Experimental investigation of stereolithography and digital light processing additive manufactured pallets

Additive manufacturing is a layer-by-layer addition of material formed by extracting the information from the digital computer-aided design (CAD) model. The CAD model is constructed and then is converted into slices (layers) by the slicer software and printed in slices, as each layer is built on the additive bed. Various additive manufacturing techniques are employed for a variety of materials and requirements, VAT polymerization refers to a technique wherein a vat of liquid photopolymer resin is used to construct the model layer by layer. An ultraviolet (UV) light is used to cure or harden the resin where required. Other additive manufacturing technologies include material extrusion, material jetting, binder jetting, powder bed fusion, sheet lamination, and direct energy deposition. Because our research on pallets is limited to the use of plastic for specialized applications, we have limited our work to focus on stereolithography (SLA) and digital light processing (DLP), both of which are subderivatives of the VAT polymerization technique.

The first additive manufacturing process was introduced in the 1980s, utilizing the ability of ultraviolet (UV) and laser light sources to cure the polymers. This process works on the vat polymerization technology with raw material about photo-curable resin. SLA, widely considered to be an accurate process for building 3D objects, has its limitations, because the build process and the laser source are point-based methods, which makes SLA slower in terms of printing time. With the advent of additive manufacturing, researchers’ interest SLA has been elevated because of the customized solution requirements, with the effect of the time factor receding. Anil et al. [12] attempted 3D printed pallets, which applied a cross-hatched design to increase the capacity of the shear factor owing to the length of the span in correlation with rectilinear fill. Researchers [13] provided a review study on stereolithography, in which medical imagery and complex surgery procedures could be enhanced with stereolithography-printed biomedical devices. With further enhancement, stereolithography was used for intravesical drug delivery in terms of flexible printed parts leading to good blood compatibility [14]. The constitution of an elastoplastic model [15] described the mechanical behaviour of the SLA-printed parts, forming initial study approach for load carriers and storage shelves in the mechanical industrial field. Furthermore, another study made use of numerical modelling to predict the elastic properties of green SLA parts of an alumina suspension for structural applications [16]. With disparate applications from wide domains, SLA-type additive manufacturing bolsters provide case-to-case solutions, especially in printing pallets for specialized packaging and logistics operations.

The process is governed by curing the raw material with UV or laser light, and the major standout in correlation with other VAT polymerization techniques lies in the fact that the DLP process involves area scanning, which tends to speed up the process and thereby reduces the time to build parts. The raw material for DLP is a photo-curable resin which acquires the required shape based on the input CAD geometry. Once the CAD geometry is converted to 2D slices in the slicing software, this process projects the 2D slice on the resin like a light projected on a projector screen and cures it based on the required parameters [17]. This process cures the resin using the pixel configuration to exploit the geometric details of the CAD part. With the modern customized solution requirements, this process is attracting much interest among researchers. It was depicted that the curing behaviour of the photopolymers availing DLP profoundly affected the reduction of the secondary UV process and enhanced the hardness and strength of the final part [18]. With advancements, DLP has been applied in the medical field in the process of creating bone implants [17], as well as in formation of self-healing hydrogels [18] attenuating strength by 72%, soft robotics (grippers) [19], energy storage [20], plastic moulds, and pallet making [20, 21].

Although both processes are classified under VAT polymerization, a thorough distinction exists between them, with layer curing either in the presence of an external light source or laser [12, 17]. With the prevalent condition of area scanning (pixel configurations), as depicted in Figure 1, the DLP process can complete the build process faster than the point-based method in SLA.

Fig. 1.

With pixel (voxel) based configuration, the accuracy and the complex profile would be compensated for to a large extent under DLP conditions as compared with point-based curing in SLA. Another beneficial aspect of DLP over SLA is the minimum pixel size as depicted in Figure 1 as correlated with the built product as showcased in Figure 2 and its correspondence to the CAD file.

Fig. 2.

Given the advancements occurring in the area of 3D printing in correlation with VAT polymerization techniques, a comprehensive evaluation of pallet formation must be performed for specialized transportation of high-risk goods: defence sectors; modular closing of aerospace parts during transportation; support structures for long, slender ratio parts as was done for the age-old enervation models made of wood (moisture absorption; reaction with chemicals; environmentally susceptible); aluminium (cost factor surge, heavy load factor of the pallet alone, mediocre impact strength) fall short for complex parts and delicate sectors. There has been limited work carried out in the area, so the authors in this work showcase the modelling of pallets to defined standards, using both SLA and DLP models. The capabilities for each are to be evaluated along with the performance of their basic simulation, through mechanical tests and surface analysis in comparison with each other. The parameters for performing various physical tests and evaluating their chemical attributes, coupons are built using layer thickness and alternative materials as parameters corresponding to tensile strength, elongation, and hardness corresponding to the levels of experiments for each VAT polymerization process. The common vector plane response variables were analysed using a hybrid AHP-TOPSIS (analytic hierarchy process and technique for order of preference by similarity to ideal solution) statistical method for individual methods to reach a solution as close as possible to the ideal solution in the experimental runs conducted. This run is utilized to print a prototype of a four-way pallet based on ISO Standard MH-2016-1 on either of the processes. Simulations are also carried out to check for the material’s stability under disparate loading conditions encountered in the industrial setting— additionally, morphological characteristics should orient the surface characteristics in regard to our applications.

The range of types of pallets often employed in the industry include four-way pallets, block pallets, stinger pallets, panel decks, and so on. The idea is to check initially for how capable the process is regarding mechanical compatibility. As aforementioned, the processes considered for checking the compatibility in additive manufacturing are SLA and DLP, for which we have considered two process parameters: layer thickness and disparate proprietary materials. For this iteration, we will employ an L9 orthogonal array to develop experimental runs. The 3D printing is based on these experimental runs, and mechanical testing is performed to determine the response variables. The response variables considered were tensile strength, percentage of elongation, and shore hardness. The AHP-TOPSIS method optimizes these response variables to develop close-to-ideal solutions in the experimental runs. Having a close-to-ideal solution, a model comprising a four-way pallet is 3D printed, using DLP and FDM, based on ISO standards.

Form labs SLA printer [21] was employed for printing the pallet samples on the basis of tensile test requirements for mechanical analysis. Table 1 depicts the process parameter values for building the test coupons.

The process parameters under consideration are layer thickness, which are 25μ, 50μ, and 100μ, and co-polymer inks, which are industrial-grade Grey V4, Clear V2, and Model V3. These materials are bio-compatible, nonreactive and toxic, thereby holding strong to the application of transporting high-value chemical goods without any surge effect due to spillage/leaks or disturbances during transport. The industrial grade material properties are listed in Table 2.

To ensure that uniformity is obtained in the analysis method and also to keep the cost factor of printing low, a L₉ orthogonal array experimental design was considered for process parameters [22–24] as indicated in Table 3, and the number of repetitions for each sample is five.

The DLP method uses the area scanning method, for which a standalone [25] 3D system printer was availed for printing of pallets and coupons related to pallet making. The process parameters for DLP 3D printing of pallets are layer thicknesses of 30μ, 40μ, and 50μ, and industry standard materials, comprising FlexBlack-20, ToughBlack-20, and ProBlack-10, as indicated in Table 4.

The industry standard material properties used for 3D printing of pallets using DLP are as indicated in Table 5.

To ensure uniformity, overall rapid analysis method, and also to keep the cost factor of printing low, an L₉ orthogonal array design of experiments was considered for process parameters as indicated in Table 6, and the number of repetitions for each sample is five in numbers [21–23] corresponding to the experimental work.

Static simulations are also carried out to check for the material’s ability in disparate loading conditions in the industrial regime. Furthermore, morphological characterization comprising SEM (surface structural attributes), AFM (waviness factor), and XRD (residual stress factor) is carried out in correlation with each printing condition to evaluate the entire model of pallet application.

To analyse the tensile mechanical characteristics of the printed parts, coupons were designed based on ASTM D638, as per standard for plastic materials. The major drawback pertaining to additive manufacturing was lack of specific standards for comparison, therefore, ASTM D638 as depicted in Figure 3, was considered for benchmarking.

Fig. 3.

The type IV CAD model was developed using the Creo tool and was formatted into a .stl file for 2D slicing software (perform for SLA; and 3D sprint for DLP), for optimal guidance and assistance (GUI) for carrying out the printing operation as depicted in Figure 4.

Fig. 4.

As per the requirements, the tensile coupons for both SLA and DLP are 3D printed as per the ASTM standard, using the machines as indicated in Figure 5. The 3D printing machine’s salient features are detailed in Table 7.

Fig. 5.

For DLP printing, the GUI interface verification adds a constructive nature, as it is a bottom-up approach, that is, material upside down printing pertaining to constraint of support structures, as depicted in Figure 6.

Fig. 6.

The pallet tensile coupons are printed in accordance with L₉ orthogonal array experiment as listed in Table 3 and Table 6 for SLA and DLP, respectively.

In commercial applications of load lifting or transportation application, pallets are manufactured, using injection moulding, which falls under the area of subtractive manufacturing. Though these pallets are used for mass production any specialized applications introduce a limiting factor, especially those that involve biomedical samples delicate aero components, thus causing a major impact from product design development to the lead time effect as the entire design and manufacturing process takes a hit. Making a die for production in large numbers also involves steps in subtractive methods, starting from rough milling, CNC machining, jig boring, die sink EDM, surface grinding, and assembly components. This eliminates specialized applications requiring an alternative method, especially for wide-scale transportation issues, as the focus lies on pallets. To recreate a real-time model, a benchmark comparison is carried out to study the mechanical characteristics of general injection moulded pallet material acrylonitrile butadiene styrene (ABS) in correlation to SLA and DLP printed pallets. The complete required die, and the manufacturing of a pallet according to an ASTM tensile coupon specimen as per commercial replicator is carried out in Central Institute of Tool Design (CITD), Hyderabad, by two-plate hand mould die with a single cavity. It is possible to execute the whole setup of making a die, to completion by injection moulding is carried out in the facilities available in CITD, Hyderabad adhering to industrial requirements as depicted in Figure 8.

Fig. 7.

Fig. 8.

The uniaxial tensile test is a destructive process that indicates information about the yield and tensile strength and measures the forces that would cause a plastic specimen to break. ASTM D638 type IV specimen 3D printed pallet specimen coupons were evaluated, using Instron model 5966, as depicted in Figure 9. The range of the machine is 0–10 kN, having a minimum and maximum speed of 0.001 mm/min and 1500 mm/min, respectively. The requisite pallet coupon specimens are placed in a universal testing machine with the aid of grippers, with a gradual load being applied to the specimens until the failure or snap region is reached. This point is dependent on the polymer conditions for which the resultant tensile loads are recorded for individual 3D printed specimens printed under SLA and DLP conditions.

Fig. 9.

In order to ensure precise uniaxial testing, specimens were marked with vertical gauges for clamping. To avoid slack from the load string for the clamping load, preload of 2N was provided to avoid uncertainty in a testing factor. The factors considered for conducting the test were approximately a load cell of 10 kN and a load rate of 5 mm/min [26, 27]. Upon breakage or snap of the polymer material, the test is automatically completed, and the results corresponding to each orthogonal array experiment for SLA and DLP were recorded accordingly. For polymer materials, the notable observation would be in the variation of tensile strength and the percentage of elongation at the breakage region and these have been tabulated in Table 8 and Table 9 for the pallet specimens corresponding to SLA and DLP, respectively.

The pallets undergo heavy loading, and thereby the hardness of the pallet specimens plays a key role in determining the load factor because the material should possess high toughness for loadcarrying capacity but should not be too brittle, which would lead to fractured, uncontrollable failure. For polymers, a shore D hardness scale durometer was used as indicated in Figure 9, as per the ASTM D2240 standard to measure the resistance to permanent deformation. The dial durometer involved pressing the indenter into the surface to capture the hardness value of the specimen at five different points, for which the average reading was tabulated for SLA and DLP pallet specimens, and the measurement has been tabulated in Table 8 and Table 9, respectively.

With preliminary experiments conducted for SLA and DLP, a common vector plane for a close to ideal solution must be obtained for correlation between values. For this, multicriteria decisionmaking methods (MCDM) were adopted, which provided decrement in statistical variability and close to a common vector plane through decision ranking. The methods VIKOR, WSM, TOPSIS, ELECTRE, DEMATEAL, MOORA, and others form different techniques that are widely adopted for solving and finding the Euclidean distance from the common vector plane [22–24, 26, 27]. For comparing and inferring a statistically close to ideal vector value, SLA and DLP, a hybrid AHP-TOPSIS methodology was adopted to find the set of optimized parameters.

The analytical hierarchy process (AHP) involves the decision-maker’s expertise to develop quantitative judgements on multicriteria evaluation. The process is based on a priority basis, categorizing attributes based on their 1 to 9 scale [30]. The multiple attributes for a complex problem are broken down into systematic hierarchical procedures to ensure low complexity and faster solutions for attributes and responses [28]. In our study, AHP was employed as a pairwise matrix comparison to come up with the weights computation, which is then utilized as a weighing criterion for response variables to optimize for closeness to the ideal solution. The AHP table formulated for our three response variables corresponding to pairwise and weights are tabulated in Tables 10 and 11.

After computing the weights for the response variables, the method was evaluated for consistency ratio. The consistency ratio was 0.081, less than 0.1, satisfying the AHP pairwise matrix usage criteria [30]. Thus, the current sets of weights were employed for hybrid analysis with the statistical TOPSIS method. For linear consistency, the same weights for data correspond to 3D-printed pallets for SLA and DLP methods, respectively.

The TOPSIS method computes the solution for the closeness to the ideal solution. The technique for order preference by similarity to the ideal solution (TOPSIS) caters for the best alternative, which has the shortest Euclidean distance from the ideal solution and is farthest from the negative ideal solution [29]. This method adopts normalising the attributes for uniformity in computing the Euclidean distances for closeness to the ideal solution. The steps involved in TOPSIS are as follows:

Step 1: The decision matrix is disguised based on the response variables.

The hybrid AHP-TOPSIS evaluation is tabulated in Tables 12 and 13 for SLA and DLP, respectively.

Based on the adaption of the hybrid method AHP-TOPSIS, the corresponding close-to-ideal values for SLA were 50μ layer thickness and Grey V4 material correspondence. Similarly, the most optimised alternative close to the ideal solution for DLP was 50μ layer thickness and ProBLK-10 material. The optimised vector plane for 3D printed pallets in their respective processes is tabulated in Table 14.

In order to produce optimised alternatives for both processes, the pallet prototypes were printed with a reduced scale of 1:20 in accordance with the ISO standard (MH1-2016), as depicted in Figure 10. The printing process is highlighted in Figure 11.

Fig. 10.

Fig. 11.

Structural analysis is needed in order to evaluate the strength of the decks and compressive strength of the geometries corresponding to pallets and to estimate the maximum loading capacity the pallets can undertake for a standard set of weights, as its main objective will be in the transportation sector.

Analysis is carried out in static structural analysis mode with one end fixed in ABAQUS software [31]. The tensile specimen is meshed with second-order reduced integration elements (C3D20R). Stress analysis is carried out on the tensile test specimen with displacement load acting on the other end of the specimen, as per the ASTM D638 standard on Inspiron-5966 in correlation with ABS (commercial replication: injection moulding), Pro-BLK-10 (SLA process), and GrayV4 (DLP method) as depicted in Figure 12.

Fig. 12.

Input data of materials stress–strain curve is considered from physical data testing after converting true stress–strain data. The displacement and von mises stress values of tensile specimens for injection-moulded and 3D printed materials are evaluated and depicted in Figure 13. The specimen behaviour is observed in each of the three cases to comprehend the results obtained in the experimental method and to evaluate the maximum stresses and displacement factor corresponding to each case for the 3D printed materials obtained.

Fig. 13.

An experimental uniaxial tensile test was carried out for 3D-printed (SLA and DLP) pallet samples in comparison with a commercial replicator for a close to ideal solution as indicated in Figure 14.

Fig. 14.

Gauge length has a colossal stress value when compared to the other regions of the specimen as a result of plastic flow or slippage. Because of the applied load and work done by an external displacement load acting on the specimen, the stress is induced in the specimen after the load application, as indicated in Figure 15, for homogenous object consideration with a constant cross-section and a constant applied load, the total deflection of the object can be determined in terms of P, L, A, and E.

Fig. 15.

The detailed comparison of simulated versus physical testing for disparate materials for each process are obtained and tabulated in Table 15.

The analytical equation is applied for the tensile specimens as per ASTM standards for given material characteristics of ABS (commercial replication: injection moulding), Pro-BLK-10 (SLA process), and GrayV4 (DLP method). The dimensional change in length is drawn from uniaxial tensile test data for each case and estimated the ratio of change length to original length for strain relation under the region of ultimate tensile strength values. Table 15 and Figure 16 show that the simulation results are ascertained closer to testing and analytical values.

Fig. 16.

Furthermore, results on the additive manufactured ISO pallet and optimized pallet conditions obtained for each process were also obtained for different loading conditions in the form of rack, fork-lift, floor, and conveyer mode, corresponding to industry applicability.

Pallets are generally classified into different categories like block types, stringer pallets, and other models based on usage criteria. Most pallets experience compressive load in practical usage and transportation. The common major subparts of pallets consist of the top deck, blocks, and bottom deck components. These subparts’ quantity and layout will depend on the size and amount of load transfer. Figure 17 indicates a generic pallet with a full uniform load spread across the pallet.

Fig. 17.

This work considers coupon levels for deriving the analytical results for displacement and stress values for pallets. The top deck is the part immediately under the load applied to the pallet, and it is a good choice for finite element analysis using beam elements for a coupon pallet size of 100 mm × 30 mm cross-section using 1D beam elements with five elements and six nodes for a 50-kg load.

In the matrix displacement method, the displacements are considered the unknowns. Each element is represented by its stiffness matrix, which relates the element-end displacements to the element-end forces for the corresponding element. The stiffness matrix for the beam elements representation is provided in Equation (2). 2 F=K∗U \[\text{F}=\text{K}*\text{U}\] where F is the local elemental forces; K represents local element stiffness metrics; and U is local displacement of the element. The elemental stiffness matrix of the beam element is indicated in Equation (3). 3 K=EIL3[ 126L−126L6L4L2−6L2L2−12−6L12−6L6L2L2−6L4L2 ] \[K=\frac{EI}{L3}\left[ \begin{matrix} 12 & 6L & -12 & 6L \\ 6L & 4{{L}^{2}} & -6L & 2{{L}^{2}} \\ -12 & -6L & 12 & -6L \\ 6L & 2{{L}^{2}} & -6L & 4{{L}^{2}} \\ \end{matrix} \right]\]

The local stiffness matrix for each of the elements is calculated and then summed up into a global stiffness matrix, for which computation is carried out by MATLAB, as represented in Figure 18.

Fig. 18.

Similarly, the force and displacement matrix is computed with similar iterations. The maximum displacement from the analytical solution is 7.8 mm at the centre of the beam location, for which corresponding 3D FEA simulations are carried out as indicated in Figure 19 and Table 16.

Fig. 19.

The 3D simulation results for 1/10 of the pallet (coupons) are interlinked closely with 1D analytical calculations using beam element with five elements. The ISO Pallet MH1-2016 and optimized conditions (close to ideal solution) are considered for analysing simulations corresponding to DLP, SLA, and ABS materials.

Von mises stress analysis is a widely used engineering tool for predicting material failure under complex loading conditions. In the case of pallet design, Von Mises stress analysis helps identify areas in a pallet where high stresses may exceed the material’s yield strength, leading to deformation or failure. Thus, it is imperative to ensure the pallet can withstand the stresses and loads it will face during use [12].

The red region on the von mises stress map indicates areas of high-stress concentration in the pallet. It can be caused by uneven loading, poor design, or material defects. By identifying these areas, designers can modify the pallet design to improve its strength and durability, such as adding reinforcement or altering the shape to reduce stress concentrations [9, 12–14].

The application of Von Mises stress analysis can be highly beneficial in pallet design to enhance the pallet’s performance, safety, and longevity. Careful consideration of high-stress areas on the stress map is a critical step in the design process. By optimizing the design of the pallet using von mises stress analysis, the risk of material failure and the associated economic and environmental impacts can be reduced. Therefore, incorporating von mises stress analysis into the pallet design process is a crucial aspect that must be considered [12, 15].

The displacement contour is a visual representation of the deformation that occurs in a structure when subjected to external loads. This tool is widely used to analyse the structural response of a component or system under varying loading conditions and helps identify areas of high-stress concentration and potential failure [17].

The displacement contour is generated through a finite element analysis process, which involves dividing the structure into a mesh of discrete elements and applying elasticity equations to calculate the deformation of each element in response to the load. The resulting displacement is plotted at each point in the structure to generate the displacement contour [12–15].

The information the displacement contour provides is crucial for optimizing the structure’s design to improve its performance and prevent failure. For example, engineers can identify areas of the structure that experience high-stress concentrations and modify the shape, size, or material properties accordingly [9, 12, 15].

ABAQUS simulates real-life loading situations to understand the material strength and loadwithstanding capacity. The disparate loading conditions availed in the study of the designed pallets are uniformly distributed flexible loads, rigid loads, rigid line loads, point loads, and discrete loads, a range that allows simulation of a real-time situation on the actual weight-carrying pallets. For the analysis of designed pallets, a uniformly distributed load of 1000 kg is used for this range of supports.

The boundary conditions used in analysing a pallet design are selected, representing floor, forklift, rack, and conveyor support, as indicated in Figure 20. The pallet meshes with first-order reduced integration elements (C3D8R) for the static simulation, as indicated in Figure 21.

Fig. 20.

Fig. 21.

Deflection and von mises stress values for the ISO pallet standard are analysed and the values are depicted in Figures 22, 23, and 24 for ABS, SLA, and DLP material, respectively.

Fig. 22.

Fig. 23.

Fig. 24.

It has been observed that the deflections and von mises stress values for the ISO pallet standard are lowest in floor support in comparison with rack, forklift, and conveyor supports, as indicated in Figure 25.

Fig. 25.

The von mises stress was analysed for all the cases subjected to pallet simulation for the ISO model. The red region, as depicted, indicates a vast amount of loading factor and maximum displacement region when the load is applied to the pallet [12].

With an increase in displacement and stress values in rack support in comparison to floor, forklift, and conveyor, a detailed comparison of the same is carried out separately for DLP and SLA, as tabulated in Table 17, Figure 26 and Table 18, and Figure 27, respectively, indicating a lower deflection in DLP condition, owing to area scanning parameters, thereby having an inherent toughness value [12, 16].

Fig. 26.

Fig. 27.

A simulation is carried out for optimized design conditions with a lower material reduction, for real-life loading situations in order to understand the material strength and load-withstanding capacity. The loading conditions used in this study are uniformly distributed flexible loads, uniformly distributed rigid loads, rigid line loads, point loads, and discrete loads on the actual product. For the analysis of designed pallets, a uniformly distributed load of 1000 kg is used for disparate supports.

The boundary conditions used in analysing a pallet design are selected, similar to the ISO pallet design, and corresponding to the floor, forklift, rack, and conveyor support, as indicated in Figure 20, but with a wide reduction in material consumption, thereby increasing the economic factor. The pallet is meshed with first-order reduced integration elements (C3D8R) for the static simulation, as indicated in Figure 28.

Fig. 28.

Deflection and von mises stress values for optimized pallet condition are analysed and the values are depicted in Figures 29, 30, and 31 for ABS, SLA, and DLP material respectively.

Fig. 29.

Fig. 30.

Fig. 31.

It has been observed that the deflections and von mises stress values for optimized pallet conditions are at their highest in floor followed by forklift support in comparison with rack and conveyor supports, as indicated in Figure 32.

Fig. 32.

The von mises stress was analysed for all cases of the optimized pallet, and the red region in the figures indicates where the highest load is experienced during uniform loading. This analysis provides engineers with valuable insights into the structural response of the pallet to applied loads and helps identify potential failure areas, which can inform design modifications to improve performance and prevent failure [12, 15].

With an increase in deflection and stress values in rack support in comparison to floor, forklift, and conveyor, a detailed comparison of the same is carried out separately for DLP and SLA processes for optimized conditions, as tabulated in Table 19, Figure 33, and Table 20, Figure 34, respectively, indicating lower deflection and stress contours in the DLP condition, owing to the areascanning parameter, thereby having inherent toughness value [12, 16].

Fig. 33.

Fig. 34.