# A Review of Process-Parameter Optimization in Fused Deposition Modeling: Effects on the Mechanical Properties of Polymer Parts

**Odilbek Marimov¹**
¹ Department of Mechanical Engineering, New Uzbekistan University, Tashkent, Uzbekistan
*Corresponding author: odilbekmarimov70@gmail.com*

> **STATUS: WORKING DRAFT — not yet submission-ready.** 

---

## Abstract

Fused deposition modeling (FDM) is the most widely adopted material-extrusion additive-manufacturing process, valued for its low cost, material range, and accessibility. However, the mechanical performance of FDM parts is highly sensitive to a large set of interacting process parameters, which complicates the production of load-bearing components. This paper reviews the published literature on the influence of the principal FDM process parameters—layer thickness, infill density and pattern, build and raster orientation, raster angle and width, air gap, extrusion (nozzle) temperature, build-plate temperature, and print speed—on tensile, flexural, and impact properties of thermoplastic parts, with emphasis on PLA, ABS, and reinforced variants. The optimization methodologies most commonly applied are then surveyed, including the Taguchi method, response surface methodology (RSM), analysis of variance (ANOVA), grey relational analysis (GRA), metaheuristic algorithms, and emerging machine-learning approaches. A comparative synthesis indicates that layer thickness and infill density are consistently among the most influential parameters for tensile strength, while build orientation governs anisotropy. The review concludes by identifying gaps—limited multi-material and multi-objective studies, inconsistent test standards, and under-use of data-driven optimization—and outlines directions for future work.

**Keywords:** additive manufacturing; fused deposition modeling; process parameters; mechanical properties; Taguchi method; optimization

---

## 1. Introduction

Additive manufacturing (AM) builds parts layer by layer directly from a digital model, enabling geometric freedom that is difficult or impossible with subtractive methods. Among AM processes, fused deposition modeling (FDM)—also termed fused filament fabrication (FFF)—is the most widespread because of its low equipment and material cost, simple operation, and broad thermoplastic compatibility *(Author, Year)*.

Despite its popularity, the adoption of FDM for functional, load-bearing parts is constrained by mechanical performance that is markedly lower and more variable than that of injection-molded equivalents. This shortfall arises from the process physics: parts are built from extruded roads that bond imperfectly, leaving inter-road and inter-layer interfaces, internal voids, and a layered, anisotropic structure. The resulting properties depend strongly on a large number of user-selectable process parameters and their interactions.

Because the parameter space is large and the responses are coupled, trial-and-error tuning is inefficient. Consequently, a substantial body of work has applied formal design-of-experiments (DOE) and optimization techniques to identify parameter settings that maximize mechanical performance. Several comprehensive reviews have summarized aspects of this effort *(e.g., Dezaki et al. / "Optimisation of Strength Properties of FDM Printed Parts—A Critical Review", 2021)*, yet the field continues to grow rapidly and results across studies are not always consistent because of differences in material, machine, specimen standard, and reported response.

This paper reviews the influence of the principal FDM process parameters on the mechanical properties of polymer parts and surveys the optimization methodologies used to tune them. The scope is limited to mechanical responses (tensile, flexural, and impact behavior) of thermoplastic FDM parts. Section 2 defines the FDM process and its key parameters. Section 3 reviews the reported effect of each parameter. Section 4 surveys optimization methodologies. Section 5 synthesizes the findings comparatively. Section 6 identifies gaps and future directions, and Section 7 concludes.

## 2. The FDM Process and Its Principal Parameters

In FDM, a thermoplastic filament is fed into a heated liquefier, extruded through a nozzle, and deposited as roads that fuse to neighboring roads and the previously deposited layer before solidifying. The quality of inter-road and inter-layer bonding, the void content, and the orientation of the roads relative to loading together determine mechanical performance.

The principal process parameters reviewed here are:

- **Layer thickness (layer height):** the height of each deposited layer.
- **Infill density:** the percentage of the internal volume filled with material.
- **Infill pattern:** the geometric pattern of the internal structure (e.g., grid, triangular, honeycomb, gyroid).
- **Build orientation:** the orientation of the part on the build plate (flat / on-edge / upright), which sets the layer direction relative to load.
- **Raster (deposition) angle:** the angle of the deposited roads relative to a reference axis (e.g., 0°, 45°, 90°).
- **Raster width and air gap:** the road width and the gap (positive or negative) between adjacent roads.
- **Extrusion / nozzle temperature:** the liquefier temperature, governing melt flow and inter-road diffusion.
- **Build-plate temperature:** affects adhesion and residual stress/warping.
- **Print speed:** the deposition speed, affecting bonding time and thermal history.
- **Number of contours / shells:** perimeters surrounding the infill.

## 3. Effects of Individual Parameters on Mechanical Properties

### 3.1 Layer thickness
Layer thickness is repeatedly reported as one of the most influential parameters for tensile strength. Thinner layers increase the number of inter-layer interfaces but generally improve interlayer bonding and reduce void content, and several studies report a negative correlation between layer thickness and ultimate tensile strength (UTS)—i.e., lower layer height yields higher UTS *(Author, Year)*. The effect is material- and machine-dependent, however, and some studies report optimal intermediate values, motivating the use of DOE rather than monotonic assumptions.

### 3.2 Infill density and pattern
Infill density correlates positively and strongly with tensile and flexural strength: higher density places more load-bearing material in the cross-section, with fully dense (100%) specimens approaching the bulk material strength *(Author, Year)*. Infill pattern also matters; comparative studies report that grid and triangular patterns tend to yield higher tensile strength than some alternative patterns, although the ranking depends on loading mode and density. Pattern selection therefore trades strength against print time and mass.

### 3.3 Build orientation and raster angle
Build orientation is the dominant source of anisotropy in FDM parts because it sets the direction of the inter-layer interfaces relative to the applied load. Specimens loaded along the roads (e.g., flat, 0° raster aligned with load) are typically far stronger than those loaded across the layers (upright), where failure occurs by inter-layer delamination *(Author, Year)*. Optimal orientations near 0°–15° are frequently reported for tensile strength, while the best orientation can differ for flexural or impact loading. Raster angle interacts with orientation; aligning rasters with the principal load direction generally maximizes strength.

### 3.4 Extrusion and build-plate temperature
Higher nozzle temperatures promote polymer inter-diffusion across road interfaces, improving bond strength up to a limit beyond which degradation or dimensional defects appear *(Author, Year)*. Build-plate temperature influences adhesion and residual stresses; inadequate temperatures can cause warping that degrades both accuracy and strength.

### 3.5 Print speed
Print speed alters the thermal history available for bonding. Higher speeds reduce bonding time and can lower strength, though within typical ranges the effect is often smaller than that of layer thickness or infill density, and some studies report negligible or non-monotonic influence *(Author, Year)*.

## 4. Optimization Methodologies

### 4.1 Taguchi method
The Taguchi method, using orthogonal arrays and signal-to-noise (S/N) ratios, is the most common approach because it efficiently screens several parameters with relatively few experiments. Numerous studies apply L9, L18, or L27 arrays to identify influential parameters and near-optimal settings for PLA, ABS, and reinforced filaments *(Author, Year)*.

### 4.2 ANOVA
Analysis of variance is routinely paired with Taguchi or factorial designs to quantify the percentage contribution and statistical significance of each parameter, providing the basis for the "most influential parameter" claims throughout the literature.

### 4.3 Response surface methodology (RSM)
RSM (often with central composite designs) models continuous parameter–response relationships and captures interactions and curvature that Taguchi screening may miss, enabling interpolation to optima between tested levels *(Author, Year)*.

### 4.4 Multi-objective and hybrid methods
Because strength, stiffness, surface finish, build time, and cost conflict, multi-objective methods are increasingly used: grey relational analysis (GRA), TOPSIS, and grey–Taguchi hybrids convert multiple responses into a single optimization index *(Author, Year)*. Metaheuristic optimizers (e.g., genetic algorithms, grey wolf optimization) are combined with RSM surrogates to search the parameter space.

### 4.5 Machine-learning approaches
A growing body of work trains data-driven models (regression, tree ensembles, neural networks) to predict mechanical properties from process parameters and to support optimization, reducing the experimental burden. This direction remains comparatively under-explored and is a clear opportunity for future research.

## 5. Comparative Synthesis

Table 1 summarizes representative studies, their parameters, materials, methods, and principal findings. *(Populate and expand this table with every study you read; aim for 20–30 rows. The rows below are starting examples drawn from the starter references.)*

**Table 1. Representative studies on FDM process-parameter optimization for mechanical properties.**

| Study (ref.) | Material | Parameters studied | Method | Key finding |
|---|---|---|---|---|
| Critical Review, Polymers 2021 [1] | PLA/ABS (review) | layer thickness, infill, orientation, raster | Literature synthesis | Layer thickness, infill density, and orientation dominate strength |
| PLA tensile study [2] | PLA | speed, infill, layer thickness, width | DOE | Highest UTS ≈ 47.8 MPa at 100% infill, 0.4 mm layer |
| PLA+ Taguchi study [3] | PLA+ | infill structure, occupancy rate, orientation | Taguchi | Tensile/impact rise with occupancy (infill) rate |
| Grey–Taguchi biopolymer [4] | Biopolymer | multiple | Grey–Taguchi | Multi-objective optimum balances tensile and modulus |
| ABS CCD/ANOVA [5] | ABS | layer thickness, orientation, raster angle/width, air gap | RSM (CCD) + ANOVA | Max tensile ≈ 50.3 MPa at 0° orientation, 0.3 mm layer |
| PA12-CF RSM/GRA/GWO [6] | PA12-CF | multiple | RSM + GRA + Grey Wolf | Hybrid optimization improves multi-response performance |

Across studies, two robust conclusions emerge: (i) **layer thickness and infill density are consistently among the most influential parameters for tensile strength**, and (ii) **build orientation is the primary driver of anisotropy**. Inconsistencies between studies are largely attributable to differences in material grade, machine, specimen standard (e.g., ASTM D638 vs. ISO 527), and the specific response optimized.

## 6. Gaps and Future Directions

1. **Standardization.** Heterogeneous specimen geometries and test standards hinder cross-study comparison; wider adoption of common standards would improve comparability.
2. **Multi-objective optimization.** Most studies optimize a single response; real parts require simultaneous trade-offs among strength, stiffness, mass, surface finish, time, and cost.
3. **Data-driven methods.** Machine-learning prediction and optimization remain under-utilized relative to classical DOE and offer efficiency gains.
4. **Interaction effects and physics-based understanding.** Many studies report main effects only; interaction effects and links to underlying bonding/void mechanisms deserve more attention.
5. **Material breadth.** Reinforced and high-performance filaments (e.g., carbon-fiber-reinforced, PEEK) are less thoroughly characterized than PLA/ABS.

## 7. Conclusions

The mechanical performance of FDM parts is governed by a coupled set of process parameters, of which layer thickness, infill density, and build orientation are the most consistently influential for tensile strength, with orientation dictating anisotropy. The Taguchi method with ANOVA dominates the optimization literature, complemented by RSM, multi-objective hybrids (GRA, TOPSIS, grey–Taguchi), metaheuristics, and emerging machine-learning approaches. Future progress will benefit from standardized testing, multi-objective and interaction-aware optimization, broader material coverage, and greater use of data-driven methods. A clearer, mechanism-linked understanding of how parameters control bonding and void formation would move the field from empirical tuning toward predictive design of FDM parts.

---

## References

[1] "Optimisation of Strength Properties of FDM Printed Parts — A Critical Review," *Polymers*, vol. 13, no. 10, art. 1587, 2021. https://www.mdpi.com/2073-4360/13/10/1587

[2] "Effects of Process Parameters on Tensile Properties of 3D-Printed PLA Parts Fabricated with the FDM Method," *PMC* (PMC12300564). https://pmc.ncbi.nlm.nih.gov/articles/PMC12300564/

[3] "Taguchi Optimization of Fused Deposition Modeling Process Parameters on Mechanical Characteristics of PLA+ Filament Material," *Scientia Iranica*. https://scientiairanica.sharif.edu/article_22362.html

[4] "Multi-Objective Optimization of Fused Deposition Modeling for Mechanical Properties of Biopolymer Parts Using the Grey-Taguchi Method," *Chinese Journal of Mechanical Engineering*, 2023 (PMC9968641). https://pmc.ncbi.nlm.nih.gov/articles/PMC9968641/

[5] "Taguchi S/N and TOPSIS Based Optimization of Fused Deposition Modelling and Vapor Finishing Process for Manufacturing of ABS Plastic Parts," *Materials* (PMC7698072). https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7698072/

[6] "Parametric Optimization of FDM Process for PA12-CF Parts Using Integrated Response Surface Methodology, Grey Relational Analysis, and Grey Wolf Optimization," *PMC* (PMC11174361). https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11174361/

[7] "Application of Taguchi Method to Optimize the Parameter of Fused Deposition Modeling (FDM) Using Oil Palm Fiber Reinforced Thermoplastic Composites," *Polymers* (PMC9182676). https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9182676/

[8] M. Rizwee and D. Kumar, "Optimizing mechanical behaviour of polylactic acid (PLA) specimen in fused deposition modeling (FDM): A Taguchi approach," *Proc. IMechE Part E*, 2026. https://journals.sagepub.com/doi/10.1177/14777606241306462

---

*Drafted June 2026 as a starting manuscript*