Data-Driven Regression Model for Systematic Control of Airless Spray Coating in Shipbuilding Industry

Data-Driven Regression Model for Systematic Control of Airless Spray Coating in Shipbuilding Industry | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Data-Driven Regression Model for Systematic Control of Airless Spray Coating in Shipbuilding Industry Jinuk Kim, Kwangyeol RYU, Yeonho Cho This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8668336/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Water ballast tanks (WBTs) are exposed to severe cyclic corrosion in marine environments. The service life of a hull often depends on whether the protective coating in WBTs satisfies the dry film thickness (DFT) requirements specified in the IMO Performance Standard for Protective Coatings (PSPC). In shipyards, epoxy primers are applied mainly by manual airless spray, and thickness is difficult to control, leading to rework from insufficient thickness or material waste from excess thickness. This study develops interpretable regression models to predict and tune key airless spray outcomes under shipyard conditions. A five-factor Central Composite Design (155 runs) varied thinning ratio (THR), pump inlet gauge pressure (PRS), tip size (TIP), stand-off distance (DIST), and gun travel speed (SPD). Responses included coating flow rate (FLOW), dry film width (DFW), average DFT, and dry film roughness (DFR). Stepwise regression with analysis of variance (ANOVA) identified significant main, interaction, and quadratic effects and yielded uncoded predictive equations. Model assumptions were assessed using residual diagnostics, and predictive accuracy was evaluated using Pred. R². All five factors significantly influenced DFT, with SPD and TIP contributing most to thickness control. DFR was dominated by PRS, THR, SPD, and DIST, reflecting coupled effects of viscosity, atomization energy, and deposition distance. The final models agreed well with experiments across the design space. The equations enable forward prediction and parameter tuning for controlled multi-pass application to achieve target DFT while reducing the risk of insufficient or excess thickness and material loss. Shipbuilding Protective coatings Airless spray Design of experiments Regression model Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 1 Introduction Water ballast tanks (WBTs) are essential components of a ship, providing necessary stability and propeller immersion, particularly when the ship is in an unloaded condition [ 1 ]. Seawater is pumped into the ballast tanks when cargo is unloaded and is discharged when cargo is loaded. This process creates a corrosive environment within the ballast tanks due to the cyclic variation between wet and high humidity conditions [ 2 ]. Therefore, there is a general consensus that the economic life of the vessel depends primarily upon the corrosion rate of its ballast tanks [ 3 ]. Among all methods of protecting WBTs from the harsh environment, protective coatings are the most widely used, offering effective protection in a relatively economical manner. However, WBTs are a complex structure, with many longitudinal and transverse reinforcements with openings such as manholes, welding scallops and drain holes, which results in many areas that are hard to coat and difficult to access [ 4 ]. To ensure adequate corrosion protection, the International Maritime Organization (IMO) established the Performance Standard for Protective Coatings (PSPC), which mandates minimum requirements for WBTs coatings on all ships above 500 gross tonnage (GT) with building contracts signed on or after July 1, 2008 [ 5 ]. Under the IMO PSPC, dry film thickness (DFT) shall be measured for quality-control purposes using appropriate film thickness gauges. The nominal dry film thickness (nDFT) for epoxy-based WBTs coatings is specified as 320 µm and is evaluated in accordance with the 90/10 rule. The 90/10 rule means that 90% of all thickness measurements shall be greater than, or equal to, nDFT and none of the remaining 10% measurements shall be below 90% of nDFT. In shipyard practice, this nDFT is typically achieved through cumulative build-up over multiple passes rather than a single pass. If the coating thickness applied to the ship block, including the WBTs area, does not meet the nDFT specified by the IMO PSPC regulations during inspection by the shipowners, repainting will be required, potentially extending the production lead time. Conversely, if the coating thickness significantly exceeds the nDFT, production costs will increase due to the excessive use of coating materials. Almost all areas of ships, including WBTs, are coated using the traditional airless spray method, which is performed manually by workers to apply protective coatings. This method has the advantage of being able to spray a large amount of paint in a short time to form a coating film, but it faces the chronic issues of difficulty in predicting DFT on the surface, making it challenging to achieve a uniform coating thickness. To address this issue, it is essential to identify the key factors influencing DFT and to develop a reliable method for quantitatively predicting DFT based on these factors. In this study, a regression model was developed using the design of experiments (DoE) approach to predict DFT as a function of key spray parameters, thereby enhancing control over the airless spray coating process and supporting compliance with PSPC requirements. To further maximize the model's applicability, additional variables including flow rate of coating material (FLOW), dry film width (DFW), and dry film roughness (DFR) were incorporated, enabling comprehensive predictions of both spray and coating properties. 2 Literature review Data-driven regression has long been used to relate spray parameters to coating outcomes. Early DoE studies in thermal or air-assisted spraying demonstrated that factorial and response-surface designs can recover main effects and selected interactions, yielding usable thickness models but with limited operational scope. For example, Datta et al. [ 6 ] combined a Central Composite Design with nonlinear regression to link gas flow, powder feed, arc current, and stand-off distance to thickness, porosity, and hardness. While this established the feasibility of multi-response modeling, the inclusion of many objectives and the natural variability of plasma spraying constrained generalization and shifted attention away from thickness-focused optimization. In HVLP settings, Luangkularb et al. [ 7 ] regressed the influence of nozzle size, atomizing pressure, and spray time under largely static conditions, offering clear trends for stationary panels but providing limited guidance for dynamic, hand-held operations typical of shipyards. Likewise, Choikhrue et al. [ 8 ] modeled a rotational spray process in which the workpiece rotates under a fixed gun. Their regression captured the effects of spindle speed and spray time but rested on simplified coating conditions that do not reflect variability observed in shipyard operations, such as gun travel, overlaps, and changing stand-off. Šolić et al. [ 9 ] present a regression for electrostatic powder coating based on voltage and current. The approach uses a narrow factor set and a static setup without motion variables, and the reported evaluation focuses on in-sample fit rather than predictive uncertainty. In parallel, physics-based computational models have sought to explicitly simulate spray/film formation. Yang et al. [ 10 ] developed a CFD framework to represent phase expansion and wall impact and compared static/dynamic simulations on planar and curved surfaces with laboratory measurements. The study is useful for understanding how geometry shapes film formation; however, it does not quantify predictive accuracy in a way that supports day-to-day process decisions. Li et al. [ 11 ] reported closer quantitative alignment between simulation and measurements for airless panels (e.g., comparable mean thickness and high transfer efficiency under their test conditions), yet comprehensive statistical validation across diverse operating conditions and simplified control-oriented rules were not provided. Wu et al. [ 12 ] further analyzed curvature and path-planning effects for complex surfaces; while the visual correspondence is informative, accuracy metrics such as MAE/RMSE are generally not emphasized, and computational costs remain non-trivial. More recently, hybrid and control-oriented approaches have begun to appear. Shi et al. [ 13 ] proposed a variable-flow model with speed planning that explicitly accounts for acceleration and deceleration during gun motion, thereby reducing local over- and under-coating compared with constant-speed/flow baselines. This highlights the importance of motion dynamics in thickness uniformity. Prior studies share several limitations that constrain practical deployment. Factor spaces were defined narrowly, and motion dynamics and cross-factor interactions were often omitted. Experimental setups were mostly simplified, leaving robustness to variability in the real operation environment uncertain. They reported how well the model fit the experimental data, not predictive statistics with quantified uncertainty, and the models appeared to work within the tested setup and provided limited evidence of transfer to other operating conditions. Moreover, they ended with mechanism description or specific case studies and did not turn the results into simple rules that can be used for real operation. This study addresses these gaps by designing a five-factor central composite design (CCD) that captures pressure, thinning ratio, tip, distance, and speed under realistic production conditions. ANOVA is applied to select statistically significant main, interaction, and quadratic effects, producing concise and easy to interpret equations. Predictive validity was quantified using Pred. R 2 and related statistics, and actionable indicators for the shipyard environment were derived to enable direct parameter tuning without heavy computation. This framework keeps regression simple and low cost, yet delivers enough accuracy for real-time process guidance in complex shipyard environments. Table 1 summarizes the differences between this study and other references. Table 1 Summary of the differences between this study and other references Study cases Technology and approach Key focus and contributions Limitations Ref. [ 6 ] Plasma spray, regression (CCD) Linked gas flow, powder feed, arc current, distance to coating outcomes Multi-response focus reduced DFT-centric optimization; plasma-specific variability Ref. [ 7 ] HVLP, regression (DoE) Effect of nozzle size, spray time, air pressure on DFT Excluded dynamic factors Ref. [ 8 ] Rotational spray, regression (DoE) Effects of nozzle rotation, spindle speed, spray time Assumes a fixed gun; limited transfer to handheld, dynamic field conditions Ref. [ 9 ] Powder coating, regression Optimization with current, voltage, substrate type Narrow variable set; excluded spray dynamics Ref. [ 10 ] Airless spray, CFD Simulated film formation, compared with experiments Computationally intensive; generalized accuracy metrics (e.g., MAE/RMSE) limited Ref. [ 11 ] Rotary airless spray, CFD-DEM Droplet collision and accumulation modeling No statistical validation; no simplified rules Ref. [ 12 ] Airless spray on spherical surface, CFD Modeled curvature and trajectory effects Did not quantify predictive accuracy for practice Ref. [ 13 ] Variable-velocity airless spray Modeled velocity-profile effects; improved uniformity vs constant baselines Improved uniformity but no closed-form predictive equations This study Airless spray, regression (CCD) Predictive equations for FLOW, DFW, DFT, DFR in shipyard coatings - 3 Methodology In this study, a DoE approach was employed to develop an airless spray model using a statistical methodology. The regression model for predicting FLOW, DFW, DFT, and DFR was derived from a dataset collected through an actual spray test with marine coating material commonly used in shipyards. The research was conducted in five stages: (1) defining the parameters and responses, (2) experimental design with DoE, (3) actual spray test with marine coating material, (4) data processing, and (5) model development. A schematic overview of the study is provided in Fig. 1 . 3.1 Definition of parameters and responses In this study, five process parameters, thinning ratio (THR), pump inlet gauge pressure (PRS), tip nozzle size (TIP), stand-off distance (DIST), and gun travel speed (SPD), were defined as system-controllable settings because they can be directly adjusted in the shipyard to regulate the operating state of airless spray coating operations. These controllable settings represent practical decision variables used by operators and supervisors to control spray property, represented by FLOW, and coating properties, represented by DFW, DFT, and DFR. 3.2 Experimental design A five-factor CCD was generated in MINITAB 17 with a two-level full-factorial core and axial augmentation (α = 2). The design comprised 80 cube points, 40 axial points, and 35 center points, resulting in a total of 155 runs in a single block. The continuous factors were studied at five coded levels (− 2, − 1, 0, + 1, +2), corresponding to the uncoded settings in Table 2 , whereas TIP was treated as a categorical factor with five levels. CCD enhances predictive reliability and enables robust estimation of curvature, thereby offering superior flexibility and precision compared to other second-order designs [ 14 ]. Table 2 Levels of the five factors in the CCD Parameters Type Subtype Coded level −2 −1 0 + 1 + 2 THR (vol.%) Numeric Continuous 4 8 12 16 20 PRS (bar) Numeric Continuous 3.00 3.75 4.50 5.25 6.00 TIP (no.) Categorical Ordinal 519 523 527 531 535 DIST (mm) Numeric Continuous 300 350 400 450 500 SPD (mm/s) Numeric Continuous 400 550 700 850 1,000 Among the studied factors, only the nozzle designation (TIP) is categorical while others are continuous variables. The TIP number is a manufacturer’s code that reflects the size of the elliptical orifice, which increases as the TIP number increases. Representative nozzle geometries for the TIP used in this work are illustrated in Fig. 2 . 3.3 Actual spray test In the spray tests, an epoxy primer commonly applied in shipyards was used and diluted with a compatible thinner to achieve the target spray viscosity. The experimental setup consisted of an airless pump, prepared substrate specimens, in-line sensors, and a robot-mounted automatic spray gun. Table 3 summarizes the key characteristics of the main components of the experimental system. Table 3 Characteristics of the main components of the experimental system Components Parameter Specification Coating material Paint product name Jotacote universal N10 Thinner product name Jotun thinner no.17 Manufacturer Jotun Solid volume ratio 72% Thinning ratio 4–20 vol.% Substrate specimen Material Carbon steel Dimension 300 mm (width) × 800 mm (height) × 2 mm (thickness) Airless pump system Pump inlet gauge pressure 3.0–6.0 bar Pressure ratio 72:1 Max pressure 432 bar Hose length 55 m Sensors Flow sensor Record real-time FLOW at 5 Hz Temperature sensor Record real-time temperature at 5 Hz Spray robot Cartesian coordinate robot Horizontal traverse at constant speed Gun Automatic airless spray For the spray experiments, the setup included an airless spray pump, a flow sensor, a temperature sensor, a Cartesian coordinate robot with an automatic spray gun, and an integrated system of hoses delivering the coating material to the spray nozzle (refer to Fig. 3 ). The robot was used to ensure repeatable stand-off distance and travel speed across runs. A flow sensor placed along the hose continuously monitored the flow rate of the coating material. During the experiments, the specimens were positioned at a consistent distance from the spray nozzle. The auto spray gun, mounted on the Cartesian robot, was programmed to move horizontally at a uniform speed while cyclically activating the spray to ensure even application. Prior to spraying, all test specimens were mechanically prepared to eliminate contaminants and surface irregularities. 3.4 Data processing The FLOW, representing the spray property, was assessed by the flow sensor located at the middle of the hoses. During each spray trial, the sensor recorded real-time flow rate data at a rate of 5 Hz as the spray gun maintained a constant travel speed, applied paint to the specimen, and subsequently stopped spraying. As demonstrated in Fig. 4 , the measured FLOW values exhibited a periodic pulsation pattern, which was attributed to the piston’s reciprocating motion within the airless pump. To ensure accuracy in average values, the FLOW data was averaged specifically in “the steady-state interval”, where the readings stabilized while the spray gun passed over the specimen and built up the coating film. The coating properties of DFW, DFT, and DFR were also individually measured. The DFW was determined by measuring the fan-pattern width (i.e., the major-axis length of the elliptical fan) using a length gauge. The upper and lower limits were visually identified at the onset boundaries of continuous film formation. DFT was measured on each coated specimen using a dry-film thickness gauge. DFT can vary widely across a surface because there is a variation in particle velocities within the fan pattern during spraying onto specimen surface [ 15 ]. Therefore, relying on single-point measurements may not yield an accurate representation of actual coating thickness [ 16 ]. Thus, it is recommended to take multiple measurements and calculate their arithmetic mean to estimate the average DFT in a given area [ 17 ]. In this study, fifty measurements of DFT were taken across the specimen surface and averaged. DFR was also measured at fifty locations using a roughness gauge with a 15 mm sampling length, and the mean value was calculated. Among the available roughness parameters, Rz was selected because it characterizes surface roughness by averaging the heights of the five highest peaks and the depths of the five deepest valleys over the sampling length, as defined by DIN 4768 [ 18 ]. In Fig. 5 , the parameter y i represents the distance between the global maximum and minimum heights measured in five consecutive sampling sections over the total sampling length. Consequently, the Rz parameter is calculated as the mean of these five y i measurements [ 19 ]. The measurement points selected for evaluating the three coating property parameters of DFW, DFT, and DFR across the specimen surface are illustrated in Fig. 6 . 3.5 Model development Using MINITAB 17 software, regression models in mathematical form were developed based on the dataset obtained from 155 experimental runs designed through CCD. Stepwise regression analysis was conducted at a significance level of 0.05 to eliminate statistically insignificant terms and to refine the overall model structure. The selection of significant predictors ensured model parsimony without compromising predictive performance. The final regression equations were validated using key statistical indicators, such as R-squared (R 2 ), adjusted R-squared (Adj. R 2 ), and predicted R-squared (Pred. R 2 ), which reflect the goodness-of-fit and the reliability of the model. 4 Results and discussion 4.1 Experimental results All experimental trials were conducted in the randomized run order generated by MINITAB 17. For brevity, Table 4 lists the first and last five runs. Analysis of variance (ANOVA) was performed to investigate how the experimental parameters affected the four response variables. A significance threshold of 0.05 was used to guide stepwise model selection, whereby statistically insignificant terms were systematically excluded. In this procedure, linear, quadratic, and two-way interaction effects among the five parameters were initially considered, and only significant predictors were retained to achieve a concise model without reducing predictive accuracy. Within the ANOVA framework, the degree of freedom (DF) represents the number of independent components available to estimate variability. The adjusted sum of squares (Adj SS) quantifies each factor’s contribution after controlling for other variables, while dividing this by the DF yields the adjusted mean square (Adj MS). The F-statistic, calculated as the ratio of Adj MS for a given factor to the mean square of the residual error, provides a test of significance. The p-value, in turn, indicates the probability of observing such an effect under the null hypothesis, with values below 0.05 confirming statistical significance. The “Error” term reflects residual variation unexplained by the model, whereas “Total” denotes the overall variability of the dataset. Taken together, these ANOVA components establish the evidential basis for term retention and provide the foundation for subsequent model evaluation. The residual error can be decomposed into lack-of-fit and pure error. Pure error reflects repeatability estimated from replicated design points, whereas lack-of-fit quantifies systematic deviation between the model and the mean response. Because the present design includes multiple center/replicated points, the lack-of-fit test can become sensitive. Therefore, model adequacy was additionally assessed using Pred. R² and residual diagnostics. To ensure field applicability, the final predictive equations were expressed in uncoded (real-world) units. Table 4 CCD runs and measured responses Test runs Parameters Responses THR (vol.%) PRS (bar) TIP (no.) DIST (mm) SPD (mm/s) FLOW (L/min) DFW (mm) DFT (µm) DFR (µm) 1 12 4.50 527 400 700 2.05 390 89.9 17.7 2 12 4.50 523 400 1,000 1.59 513 48.5 30.6 3 8 3.75 523 350 550 1.36 409 81.1 22.6 4 8 3.75 527 450 550 1.49 420 83.9 28.9 5 16 3.75 535 450 850 2.60 480 77.0 25.3 ⁝ ⁝ ⁝ ⁝ ⁝ ⁝ ⁝ ⁝ ⁝ ⁝ 151 16 3.75 523 350 550 1.57 445 82.1 20.3 152 8 5.25 527 350 850 2.04 371 90.4 22.2 153 8 3.75 535 350 550 1.72 301 145.3 23.3 154 16 5.25 531 350 850 3.19 400 106.7 22.2 155 8 5.25 535 350 850 2.44 318 120.9 24.7 4.1.1 FLOW According to ANOVA summarized in Table 5 , the regression model for FLOW was statistically significant. The main effects of TIP, PRS, and THR were all highly significant, and two interaction terms of THR × TIP and PRS × TIP also showed significant contributions. Table 5 ANOVA of FLOW response Source DF Adj SS Adj MS F -value p -value Model 16 62.4179 3.90112 553.08 < 0.001 Linear 6 57.4272 9.57121 1356.95 < 0.001 THR 1 9.5688 9.56884 1356.62 < 0.001 PRS 1 9.1246 9.12457 1293.63 < 0.001 TIP 4 38.7338 9.68346 1372.86 < 0.001 Square 1 0.0291 0.02909 4.12 0.044 THR × THR 1 0.0291 0.02909 4.12 0.044 2-Way interactions 9 4.9616 0.55128 78.16 < 0.001 THR × PRS 1 0.1253 0.12529 17.76 < 0.001 THR × TIP 4 3.3621 0.84051 119.16 < 0.001 PRS × TIP 4 1.4742 0.36855 52.25 < 0.001 Error 138 0.9734 0.00705 Lack-of-Fit 108 0.8407 0.00778 1.76 0.038 Pure Error 30 0.1327 0.00442 Total 154 63.3913 Among these, TIP was found to have the most dominant influence on FLOW, as it directly affects the cross-sectional area through which the coating material passes. Among the interaction terms, THR×TIP exhibited the largest effect, while PRS×TIP and THR×PRS also remained statistically significant. The uncoded regression equations for FLOW, corresponding to each TIP nozzle size, are summarized in Table 6 . Table 6 Uncoded regression equations for FLOW by TIP TIP Uncoded regression equation for FLOW (L/min) 519 0.880 − 0.0192·THR + 0.0060·PRS − 0.000879·THR² + 0.01319·THR·PRS 523 0.751 − 0.0035·THR + 0.0746·PRS − 0.000879·THR² + 0.01319·THR·PRS 527 0.009 + 0.0276·THR + 0.2433·PRS − 0.000879·THR² + 0.01319·THR·PRS 531 −0.469 + 0.0619·THR + 0.3278·PRS − 0.000879·THR² + 0.01319·THR·PRS 535 −0.826 + 0.0949·THR + 0.3951·PRS − 0.000879·THR² + 0.01319·THR·PRS The model exhibited the highest predictive performance among all four responses, with a Pred. R² of 0.9805, along with R² = 0.9846 and Adj. R² = 0.9829. These values indicate that the regression model is highly reliable in estimating FLOW under a wide range of operating conditions. In practical terms, FLOW increased with higher values of PRS and THR, as well as with the use of larger nozzle sizes. These relationships are intuitive, as greater pressure and lower viscosity facilitate increased spray volume of coating material. 4.1.2 DFW The regression model for DFW also demonstrated strong statistical significance, with ANOVA results shown in Table 7 . Table 7 ANOVA of DFW response Source DF Adj SS Adj MS F -value p -value Model 16 683,058 42,691 142.45 < 0.001 Linear 7 651,849 93,121 310.72 < 0.001 THR 1 48,088 48,088 160.46 < 0.001 PRS 1 1,397 1,397 4.66 0.033 DIST 1 271,758 271,758 906.78 < 0.001 TIP 4 330,606 82,652 275.78 < 0.001 2-Way interactions 9 31,209 3,468 11.57 < 0.001 THR × PRS 1 1,373 1,373 4.58 0.034 THR × TIP 4 24,260 6,065 20.24 < 0.001 DIST × TIP 4 5,576 1,394 4.65 0.001 Error 138 41,358 300 Lack-of-Fit 108 40,632 376 15.54 < 0.001 Pure Error 30 726 24 Total 154 724,416 The most influential parameter was DIST, followed by TIP, THR, and PRS. Additionally, the interaction between THR and TIP was statistically significant. The strong effect of stand-off distance on film width is consistent with the physical behavior of spray dispersion. The regression equations in uncoded units for each TIP nozzle are presented in Table 8 . Table 8 Uncoded regression equations for DFW by TIP TIP Uncoded regression equation for DFW (mm) 519 −47.8 + 6.86·THR + 21.12·PRS + 1.0812·DIST − 1.381·THR·PRS 523 −112.1 + 10.63·THR + 21.12·PRS + 1.1507·DIST − 1.381·THR·PRS 527 −51.4 + 8.79·THR + 21.12·PRS + 0.8212·DIST − 1.381·THR·PRS 531 −118.7 + 12.60·THR + 21.12·PRS + 0.8565·DIST − 1.381·THR·PRS 535 −171.6 + 17.22·THR + 21.12·PRS + 0.8493·DIST − 1.381·THR·PRS Model evaluation metrics indicated a high level of fit, with R² = 0.9429, Adj. R² = 0.9363, and Pred. R² = 0.9217. These results confirm that the regression model accurately describes the relationship between operating parameters and DFW. As expected, increased spray distance led to wider fan patterns and thus increased DFW. Conversely, larger nozzle sizes (higher TIP codes) tended to reduce DFW, which corresponds to narrower spray angles for larger nozzles. According to Lefebvre [ 20 ], increasing the nozzle orifice diameter leads to a thicker annular liquid sheet at the exit, which generates larger ligaments and droplets via breakup mechanisms. However, such thick sheets are less likely to spread laterally, as their inertia suppresses sheet widening. As a result, the spray angle narrows, and the overall fan width is reduced despite the increase in flow rate. This theoretical explanation is consistent with our empirical observation that higher TIP numbers yield narrower dry film width. 4.1.3 DFT Among the four responses, the regression model for DFT is of greatest practical importance, particularly because of its relevance to regulatory compliance under IMO PSPC. As summarized in Table 9 , all five main effects of THR, PRS, TIP, DIST, and SPD were statistically significant, with SPD and TIP exerting the strongest main effects on DFT. In addition, five two-way interaction terms also showed significant influence, with the SPD × TIP interaction being the most pronounced, indicating complex interdependencies between the operating parameters. Table 9 ANOVA of DFT response Source DF Adj SS Adj MS F -value p -value Model 23 191,100 8,308.7 221.17 < 0.001 Linear 8 178,206 22,275.8 592.96 < 0.001 THR 1 1,037 1,037.3 27.61 < 0.001 PRS 1 13,951 13,951.0 371.36 < 0.001 DIST 1 11,067 11,067.2 294.60 < 0.001 SPD 1 44,415 44,415.3 1,182.29 < 0.001 TIP 4 107,736 26,933.9 716.95 < 0.001 Square 1 2,283 2,282.7 60.76 < 0.001 SPD × SPD 1 2,283 2,282.7 60.76 < 0.001 2-Way interactions 14 10,611 757.9 20.18 < 0.001 PRS × SPD 1 453 452.7 12.05 0.001 PRS × TIP 4 3,223 805.8 21.45 < 0.001 DIST × SPD 1 526 525.8 14.00 < 0.001 DIST × TIP 4 1,295 323.7 8.62 < 0.001 SPD × TIP 4 5,115 1,278.7 34.04 < 0.001 Error 131 4,921 37.6 Lack-of-Fit 101 4,872 48.2 29.36 < 0.001 Pure Error 30 49 1.6 Total 154 196,022 The TIP-specific uncoded regression equations are listed in Table 10 . Table 10 Uncoded regression equations for DFT by TIP TIP Uncoded regression equation for DFT (µm) 519 221.9 + 0.735·THR + 20.22·PRS − 0.3434·DIST − 0.3542·SPD + 0.000175·SPD² − 0.02115·PRS·SPD + 0.000342·DIST·SPD 523 249.6 + 0.735·THR + 22.90·PRS − 0.3679·DIST − 0.3772·SPD + 0.000175·SPD² − 0.02115·PRS·SPD + 0.000342·DIST·SPD 527 309.8 + 0.735·THR + 29.77·PRS − 0.4464·DIST − 0.4243·SPD + 0.000175·SPD² − 0.02115·PRS·SPD + 0.000342·DIST·SPD 531 338.9 + 0.735·THR + 34.10·PRS − 0.4878·DIST − 0.4473·SPD + 0.000175·SPD² − 0.02115·PRS·SPD + 0.000342·DIST·SPD 535 358.6 + 0.735·THR + 38.90·PRS − 0.5113·DIST − 0.4717·SPD + 0.000175·SPD² − 0.02115·PRS·SPD + 0.000342·DIST·SPD Although the Pred. R² for DFT was 0.9620, slightly lower than that of FLOW, the model still exhibited excellent predictive capacity (R² = 0.9749, Adj. R² = 0.9705), making it a robust tool for field implementation. From a practical perspective, DFT increased with higher values of PRS and larger nozzle sizes, while it decreased with greater DIST and SPD. The magnitude of the effect from TIP and SPD was particularly large. These findings suggest that achieving the nominal dry film thickness (nDFT) requires coordinated adjustment of not only the material delivery (e.g., PRS, THR) but also the spray gun movement dynamics (e.g., SPD, DIST). Therefore, the DFT model serves as the most critical foundation for establishing field-oriented spray control strategies. 4.1.4 DFR For DFR, Table 11 shows that THR and SPD are the dominant main effects, while PRS, DIST, and TIP make smaller yet still significant contributions. The response is further shaped by three significant two-way interactions, with THR × PRS exerting the strongest influence, followed by THR × DIST and THR × SPD. Table 11 ANOVA of DFR response Source DF Adj SS Adj MS F -value p -value Model 14 4042.15 288.72 40.68 < 0.001 Linear 8 3143.05 392.88 55.35 < 0.001 THR 1 1384.34 1384.34 195.02 < 0.001 PRS 1 474.82 474.82 66.89 < 0.001 DIST 1 235.18 235.18 33.13 < 0.001 SPD 1 685.31 685.31 96.55 < 0.001 TIP 4 363.41 90.85 12.80 < 0.001 Square 1 323.68 323.68 45.60 < 0.001 THR × THR 1 323.68 323.68 45.60 < 0.001 2-Way interactions 5 575.42 115.08 16.21 < 0.001 THR × PRS 1 198.06 198.06 27.90 < 0.001 THR × DIST 1 159.14 159.14 22.42 < 0.001 THR × SPD 1 122.78 122.78 17.30 < 0.001 PRS × DIST 1 60.66 60.66 8.55 0.004 PRS × SPD 1 34.78 34.78 4.90 0.028 Error 140 993.76 7.10 Lack-of-Fit 110 986.11 8.96 35.15 < 0.001 Pure Error 30 7.65 0.26 Total 154 5035.91 The regression equations for DFR in uncoded units are summarized in Table 12 . Table 12 Uncoded regression equations for DFR by TIP TIP Uncoded regression equation for DFR (µm) 519 −47.1 − 1.168·THR + 4.44·PRS + 0.2171·DIST + 0.0671·SPD + 0.0927·THR² + 0.5245·THR·PRS − 0.00705·THR·DIST − 0.002065·THR·SPD − 0.02322·PRS·DIST − 0.00586·PRS·SPD 523 −45.9 − 1.168·THR + 4.44·PRS + 0.2171·DIST + 0.0671·SPD + 0.0927·THR² + 0.5245·THR·PRS − 0.00705·THR·DIST − 0.002065·THR·SPD − 0.02322·PRS·DIST − 0.00586·PRS·SPD 527 −49.8 − 1.168·THR + 4.44·PRS + 0.2171·DIST + 0.0671·SPD + 0.0927·THR² + 0.5245·THR·PRS − 0.00705·THR·DIST − 0.002065·THR·SPD − 0.02322·PRS·DIST − 0.00586·PRS·SPD 531 −48.4 − 1.168·THR + 4.44·PRS + 0.2171·DIST + 0.0671·SPD + 0.0927·THR² + 0.5245·THR·PRS − 0.00705·THR·DIST − 0.002065·THR·SPD − 0.02322·PRS·DIST − 0.00586·PRS·SPD 535 −45.8 − 1.168·THR + 4.44·PRS + 0.2171·DIST + 0.0671·SPD + 0.0927·THR² + 0.5245·THR·PRS − 0.00705·THR·DIST − 0.002065·THR·SPD − 0.02322·PRS·DIST − 0.00586·PRS·SPD The model for DFR exhibited acceptable performance, with R² = 0.8027, Adj. R² = 0.7829, and Pred. R² = 0.7539. While lower than other response models, the performance was sufficient to provide meaningful process insights. Interpretation of the results showed that smoother surfaces (i.e., lower DFR values) were achieved under higher PRS and THR conditions, likely due to improved atomization and better surface leveling. In contrast, faster spray speeds and longer spray distances tended to increase surface roughness, likely due to incomplete film formation and reduced droplet overlap. This trend aligns with previous studies. Luangkularb et al. [ 7 ] reported that increased spray distance reduces droplet momentum, leading to poor droplet coalescence and higher roughness. Similarly, Abe et al. [ 21 ] observed that excessive spray velocity or stand-off distance impairs splat deformation and bonding, resulting in a more porous and irregular coating surface. These findings highlight the importance of controlling application dynamics in order to improve the surface quality of topcoats or anti-fouling coatings in marine applications. Residual diagnostics (Fig. 7 ) indicated approximate normality and no major departures from model assumptions within the investigated region, with a slight increase in variance at higher fitted values. 4.2 Discussion 4.2.1 Factor Effects on Spray and Coating Properties The regression analysis quantified the influence of five operating parameters of THR, PRS, TIP, DIST, and SPD on four spray and coating properties: FLOW, DFW, DFT, and DFR. In addition to main effects, two-way interaction terms were selectively retained when statistically significant. The dominant contributing factors for each response are summarized in Table 13 . Table 13 Influential factors per response (main effects and key interactions) Response Dominant Factors (descending order of influence) FLOW TIP > PRS > THR (interactions: THR×TIP, PRS×TIP) DFW DIST > TIP > THR > PRS (interactions: THR×TIP, DIST×TIP) DFT TIP ≈ SPD > PRS > DIST > THR (interactions: TIP×SPD, DIST×SPD, PRS×TIP) DFR THR > PRS > SPD > DIST (interactions: THR×PRS, THR×DIST, THR×SPD) 4.2.2 Regression Model Performance Using a Central Composite Design (CCD) with 155 experimental runs, regression models were developed for all four responses. Stepwise selection was employed to retain only statistically significant terms. The final models demonstrated strong predictive performance as summarized in Table 14 . Table 14 Regression model performance Responses R 2 Adj. R 2 Pred. R 2 FLOW 0.9846 0.9829 0.9805 DFW 0.9429 0.9363 0.9217 DFT 0.9749 0.9705 0.9620 DFR 0.8027 0.7829 0.7539 The FLOW, DFW, and DFT models all showed excellent fit and generalization capability, with Pred. R² exceeding 92%. Although the DFR model had a lower Pred. R², the model still captures key trends and offers useful insights for managing surface quality in shipyard applications. To ensure practical usability, all regression equations were provided in uncoded units, and most were stratified by TIP. This enables direct application in shipyard environments without the need for transformation or post-processing. For a more comprehensive assessment of the proposed model’s predictive ability, its performance was benchmarked against regression-based DFT models reported in prior studies using Pred. R² as a common reference metric. As summarized in Fig. 8 , the proposed model achieved Pred. R² = 0.962 and compares favorably with the regression models of Datta et al. [ 6 ], Luangkularb et al. [ 7 ], Choikhrue et al. [ 8 ], and Šolić et al. [ 9 ]. These results support the robustness of the developed airless-spray model and indicate strong predictive reliability for industrial coating applications within the investigated operating window. Although the lack-of-fit test indicated statistically detectable deviations from the second-order form, the equations were retained to preserve interpretability for field tuning. Within the investigated CCD region, predictive utility was supported by Pred. R² and residual diagnostics (Fig. 7 ), and the equations should be used for prediction and parameter tuning within this region rather than extrapolated beyond it. 5 Conclusions This study developed and validated regression models for four key outcomes of FLOW, DFW, DFT, and DFR for the airless spray coating process in shipbuilding using a five-factor Central Composite Design and stepwise term selection. The models were fitted to experimental data acquired under realistic shipyard operation conditions; only statistically significant main, interaction, and quadratic terms were retained, yielding concise yet accurate predictive equations. The factor-importance analysis showed that DFT is predominantly governed by TIP and SPD (followed by PRS, DIST, THR), DFW is driven mainly by DIST and TIP, FLOW is most sensitive to TIP and PRS (then THR), and DFR depends strongly on THR and PRS. Moreover, several interactions also play meaningful roles. The regression models exhibited strong predictive performance: Pred. R² was 0.9805 for FLOW, 0.9217 for DFW, 0.9620 for DFT, and 0.7539 for DFR (with corresponding R² and Adj. R² values also high), indicating excellent fit and generalization for FLOW, DFW, and DFT, and practically useful trends for DFR. To facilitate direct field use, the final equations were expressed in uncoded (real-world) units and, where appropriate, stratified by nozzle TIP, enabling practitioners to set operating conditions without additional transformation or post-processing. Operators and supervisors can use the regression equations to tune process parameters, maintain target DFT, and achieve high film quality. Benchmarking against selected prior regression-based DFT studies indicates that the present DFT model shows competitive predictive performance (Pred. R² = 0.962) relative to predictive indicators reported in the literature. This result supports the robustness of the proposed approach and its practical utility for guiding coating operations within the investigated process window. Overall, the work delivers an interpretable, high-accuracy, and deployment-ready predictive framework for airless spray coating in shipbuilding, offering clear parameter guidance to achieve target DFT while balancing fan width, flow, and surface quality. Declarations Ethics approval Herewith the confirmation: The paper was and is not submitted for publication elsewhere. This paper has not been published elsewhere in its entirety, in part, or in a modified version. The paper was not submitted for possible publication elsewhere. Consent for publication Not applicable. Consent to participate Not applicable. Competing interests The authors have no relevant financial or non-financial interests to disclose. Funding The authors declare that no funds, grants, or other support were received for conducting this research. Author Contributions Jinuk Kim: Conceptualization, Data curation, Formal analysis, Investigation, Software, Writing – original draft. Kwangyeol Ryu: Methodology, Supervision, Validation, Writing – review & editing. Yeonho Cho: Project administration, Resources. All authors have read and approved the final manuscript. Data availability The data that support the findings of this study are available from the corresponding author upon reasonable request, due to proprietary agreements with the collaborating shipyard. References Oriaifo E, Perera N, Guy A, Leung PS, Tan K (2014) A review of test protocols for assessing coating performance of water ballast tank coatings. In: The ICMNE 2014: XII International Conference on Marine and Naval Engineering, Istanbul, Turkey, 29–30 September 2014. https://doi.org/10.5281/zenodo.1096381 Iannarelli P, Beaumont D, Liu Y et al (2022) The degradation mechanism of a marine coating under service conditions of water ballast tank. Prog Org Coat 162:106588. https://doi.org/10.1016/j.porgcoat.2021.106588 De Baere K, Verstraelen H, Rigo P, Van Passel S, Lenaerts S, Potters G (2013) Reducing the cost of ballast tank corrosion: an economic modeling approach. Mar Struct 32:136–152. https://doi.org/10.1016/j.marstruc.2012.10.009 Willemen R, Luyckx D, Meskens R, Lenaerts S, De Baere K (2020) A study into the coating thickness of ship ballast tanks. Int J Mar Eng 162(A3):277–287. https://doi.org/10.5750/ijme.v162iA3.1137 International Maritime Organization (IMO) (2006) Resolution MSC.215(82): Performance Standard for Protective Coatings for Dedicated Seawater Ballast Tanks in All Types of Ships and Double-Side Skin Spaces of Bulk Carriers (adopted on 8 December 2006. IMO, London. effective 1 July 2008) Datta S, Pratihar DK, Bandyopadhyay PP (2013) Modeling of plasma spray coating process using statistical regression analysis. Int J Adv Manuf Technol 65:967–980. https://doi.org/10.1007/s00170-012-4232-y Luangkularb S, Prombanpong S, Tangwarodomnukun V (2014) Material consumption and dry film thickness in spray coating process. Procedia CIRP 17:789–794. https://doi.org/10.1016/j.procir.2014.02.046 Choikhrue M, Prombanpong S, Sriyotha P (2016) An effect of coating parameters to dry film thickness in spray coating process. Key Engineering Materials 709:95–98. https://doi.org /10.4028/www.scientific.net/KEM.709.95 Šolić T, Marić D, Peko I, Samardžić I (2025) Optimization of Coating Process Parameters by Analysis of Target Powder Thickness and Regression Modeling. Appl Sci 15(2):673. https://doi.org/10.3390/app15020673 Yang G, Wu Z, Chen Y, Chen S, Jiang J (2022) Modeling and characteristics of airless spray film formation. Coatings 12(7):949. https://doi.org/10.3390/coatings12070949 Li X, Chen X, Hong N, Li Q, Xu Z, Sheng M, Wang R (2024) A CFD-DEM simulation of droplets in an airless spray coating process of a square duct. Coatings 14(3):282. https://doi.org/10.3390/coatings14030282 Wu Z, Wang C, Yang G, Chen S, Duan J, Chen Y (2024) Numerical investigation of film formation characteristics and mechanisms through airless spraying on spherical surfaces. Coatings 14(10):1299. https://doi.org/10.3390/coatings14101299 Shi T, Xu J, Cui J, Tao L, Xu W, Wang Z, Ji J (2023) Variable velocity coating thickness distribution model for super-large planar robot spraying. Coatings 13(8):1434. https://doi.org/10.3390/coatings13081434 Myers RH, Montgomery DC, Anderson-Cook CM (2016) Response surface methodology: Process and product optimization using designed experiments, 4th edn. Wiley, Hoboken Kim SK, Jang SK, Kim DK, Jang JR (2011) Design and performance evaluation of airless nozzle for painting. Proceedings of the Korean Society for Precision Engineering Spring Conference (Part II), pp 1363–1364 ASTM D7091-22 (2022) Standard Practice for Nondestructive Measurement of Dry Film Thickness of Nonmagnetic Coatings Applied to Ferrous Metals and Nonmagnetic, Nonconductive Coatings Applied to Non-Ferrous Metals. ASTM International, West Conshohocken, PA, USA SSPC-PA 2 (2022) Procedure for Determining Conformance to Dry Coating Thickness Requirements. AMPP (Association for Materials Protection and Performance), Houston, TX, USA (2022) Deutsches Institut für Normung (DIN) (1990) DIN 4768: Determination of values of surface roughness parameters Ra, Rz, Rmax using electrical contact (stylus) instruments—Concepts and measuring conditions Amaral MM, Raele MP, Caly JP, Samad RE, Vieira ND Jr, Freitas AZ (2009) Roughness measurement methodology according to DIN 4768 using optical coherence tomography (OCT). In: Modeling Aspects in Optical Metrology II. https://doi.org/10.1117/12.827748 Lefebvre AH, McDonell VG (2017) Atomization and sprays, 2nd edn. CRC, Boca Raton Abe A, Goss JA, Zou M (2024) Exploring the impact of spray process parameters on graphite coatings: Morphology, thickness, and tribological properties. Coatings 14(6):714. https://doi.org/10.3390/coatings14060714 Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8668336","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":589837614,"identity":"b1ba7c3a-7937-47ef-a30c-c2f1ff4f1154","order_by":0,"name":"Jinuk Kim","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Jinuk","middleName":"","lastName":"Kim","suffix":""},{"id":589837615,"identity":"e92ad3bf-6997-4a36-8477-d988b44dc6b9","order_by":1,"name":"Kwangyeol RYU","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAzElEQVRIiWNgGAWjYBACCTBZgRAwIFLLGZK1MLaRokVyRu4zad55dxLnt589wPCjhsHYvIGAFmmJdDNp3m3PEjecyUtg7DnGYCZzgIAWOYk0NqCWw4kbGHIMGHgbGGwkCDkMomXO4cT5/W8MGP8So0UarKXhcGLDjRwDZqAtZgS1SPY8Y7acc+yZ8YYbbwwOyxyTMCaoReJ4GuONNzV3ZOf35xg+fFNjYziDkBYGgQQWoLkHwOwDsKjFD/gPMH+AaRkFo2AUjIJRgBUAAANmO0BPjc1WAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0003-0892-5684","institution":"Pusan National University","correspondingAuthor":true,"prefix":"","firstName":"Kwangyeol","middleName":"","lastName":"RYU","suffix":""},{"id":589837616,"identity":"0e11aa81-7c15-40ac-b993-8909822f8d95","order_by":2,"name":"Yeonho Cho","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Yeonho","middleName":"","lastName":"Cho","suffix":""}],"badges":[],"createdAt":"2026-01-22 10:14:10","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8668336/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8668336/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":102965024,"identity":"f607855d-cbcd-4ef1-8a70-dbcd952dffa6","added_by":"auto","created_at":"2026-02-19 04:30:04","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":253956,"visible":true,"origin":"","legend":"\u003cp\u003eSchematic flow of study\u003c/p\u003e","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8668336/v1/61c222df5142c32dc175014d.jpeg"},{"id":102964920,"identity":"86a33592-6e21-4351-a98f-6d00fae224b7","added_by":"auto","created_at":"2026-02-19 04:29:10","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":147800,"visible":true,"origin":"","legend":"\u003cp\u003eTIP numbers and shapes\u003c/p\u003e","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8668336/v1/f6c28e5f4313bf655df42ef5.jpeg"},{"id":102966105,"identity":"80b30007-0a16-4029-87cf-31d5f1fe3ce9","added_by":"auto","created_at":"2026-02-19 04:34:46","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":91887,"visible":true,"origin":"","legend":"\u003cp\u003eThe setup of experimental system\u003c/p\u003e","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8668336/v1/0f2cf55d32d55e3b5053cb31.jpeg"},{"id":102965026,"identity":"98e6baea-63e8-43ba-8d77-7593d3ebbfaa","added_by":"auto","created_at":"2026-02-19 04:30:05","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":60877,"visible":true,"origin":"","legend":"\u003cp\u003eReal-time plot of FLOW\u003c/p\u003e","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8668336/v1/830c8cf53ce8e3e8cb3d68ee.jpeg"},{"id":102965020,"identity":"40086b1c-2280-46a4-9853-133a29e10b1c","added_by":"auto","created_at":"2026-02-19 04:29:59","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":146022,"visible":true,"origin":"","legend":"\u003cp\u003eRepresentation of the average depth of roughness parameter of Rz\u003c/p\u003e","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8668336/v1/6da140bc67c4ccd5efc1e167.jpeg"},{"id":102965036,"identity":"c27ec2c8-1d49-4935-8248-abe2e016020e","added_by":"auto","created_at":"2026-02-19 04:30:09","extension":"jpeg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":236609,"visible":true,"origin":"","legend":"\u003cp\u003eThe measurement locations of the three coating properties\u003c/p\u003e","description":"","filename":"floatimage6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8668336/v1/47dca21633e29a49bc193ffb.jpeg"},{"id":102965006,"identity":"e96d879d-2dc1-4673-bcb0-5234307f7099","added_by":"auto","created_at":"2026-02-19 04:29:50","extension":"jpeg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":967507,"visible":true,"origin":"","legend":"\u003cp\u003eResidual diagnostic plots for the developed regression models\u003c/p\u003e","description":"","filename":"floatimage7.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8668336/v1/b1f932810dba96c16cdcb615.jpeg"},{"id":102965208,"identity":"92e59d17-b74c-4d4f-be96-ea9849b557b5","added_by":"auto","created_at":"2026-02-19 04:30:48","extension":"jpeg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":158291,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of DFT Prediction Performance (Pred. R²)\u003c/p\u003e","description":"","filename":"floatimage8.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8668336/v1/8fb95cdafeee41152f22f6a3.jpeg"},{"id":106402336,"identity":"d1d407a2-10bf-467c-b138-8dca916ebf87","added_by":"auto","created_at":"2026-04-08 09:11:46","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3362879,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8668336/v1/64356dac-87d4-409a-886f-090d3a07f2e8.pdf"}],"financialInterests":"","formattedTitle":"Data-Driven Regression Model for Systematic Control of Airless Spray Coating in Shipbuilding Industry","fulltext":[{"header":"1 Introduction","content":"\u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eWater ballast tanks (WBTs) are essential components of a ship, providing necessary stability and propeller immersion, particularly when the ship is in an unloaded condition [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Seawater is pumped into the ballast tanks when cargo is unloaded and is discharged when cargo is loaded. This process creates a corrosive environment within the ballast tanks due to the cyclic variation between wet and high humidity conditions [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Therefore, there is a general consensus that the economic life of the vessel depends primarily upon the corrosion rate of its ballast tanks [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eAmong all methods of protecting WBTs from the harsh environment, protective coatings are the most widely used, offering effective protection in a relatively economical manner. However, WBTs are a complex structure, with many longitudinal and transverse reinforcements with openings such as manholes, welding scallops and drain holes, which results in many areas that are hard to coat and difficult to access [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eTo ensure adequate corrosion protection, the International Maritime Organization (IMO) established the Performance Standard for Protective Coatings (PSPC), which mandates minimum requirements for WBTs coatings on all ships above 500 gross tonnage (GT) with building contracts signed on or after July 1, 2008 [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. Under the IMO PSPC, dry film thickness (DFT) shall be measured for quality-control purposes using appropriate film thickness gauges. The nominal dry film thickness (nDFT) for epoxy-based WBTs coatings is specified as 320 \u0026micro;m and is evaluated in accordance with the 90/10 rule. The 90/10 rule means that 90% of all thickness measurements shall be greater than, or equal to, nDFT and none of the remaining 10% measurements shall be below 90% of nDFT. In shipyard practice, this nDFT is typically achieved through cumulative build-up over multiple passes rather than a single pass. If the coating thickness applied to the ship block, including the WBTs area, does not meet the nDFT specified by the IMO PSPC regulations during inspection by the shipowners, repainting will be required, potentially extending the production lead time. Conversely, if the coating thickness significantly exceeds the nDFT, production costs will increase due to the excessive use of coating materials.\u003c/p\u003e \u003cp\u003eAlmost all areas of ships, including WBTs, are coated using the traditional airless spray method, which is performed manually by workers to apply protective coatings. This method has the advantage of being able to spray a large amount of paint in a short time to form a coating film, but it faces the chronic issues of difficulty in predicting DFT on the surface, making it challenging to achieve a uniform coating thickness. To address this issue, it is essential to identify the key factors influencing DFT and to develop a reliable method for quantitatively predicting DFT based on these factors.\u003c/p\u003e \u003cp\u003eIn this study, a regression model was developed using the design of experiments (DoE) approach to predict DFT as a function of key spray parameters, thereby enhancing control over the airless spray coating process and supporting compliance with PSPC requirements. To further maximize the model's applicability, additional variables including flow rate of coating material (FLOW), dry film width (DFW), and dry film roughness (DFR) were incorporated, enabling comprehensive predictions of both spray and coating properties.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e"},{"header":"2 Literature review","content":"\u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eData-driven regression has long been used to relate spray parameters to coating outcomes. Early DoE studies in thermal or air-assisted spraying demonstrated that factorial and response-surface designs can recover main effects and selected interactions, yielding usable thickness models but with limited operational scope. For example, Datta et al. [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e] combined a Central Composite Design with nonlinear regression to link gas flow, powder feed, arc current, and stand-off distance to thickness, porosity, and hardness. While this established the feasibility of multi-response modeling, the inclusion of many objectives and the natural variability of plasma spraying constrained generalization and shifted attention away from thickness-focused optimization. In HVLP settings, Luangkularb et al. [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e] regressed the influence of nozzle size, atomizing pressure, and spray time under largely static conditions, offering clear trends for stationary panels but providing limited guidance for dynamic, hand-held operations typical of shipyards. Likewise, Choikhrue et al. [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e] modeled a rotational spray process in which the workpiece rotates under a fixed gun. Their regression captured the effects of spindle speed and spray time but rested on simplified coating conditions that do not reflect variability observed in shipyard operations, such as gun travel, overlaps, and changing stand-off. Šolić et al. [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e] present a regression for electrostatic powder coating based on voltage and current. The approach uses a narrow factor set and a static setup without motion variables, and the reported evaluation focuses on in-sample fit rather than predictive uncertainty.\u003c/p\u003e \u003cp\u003eIn parallel, physics-based computational models have sought to explicitly simulate spray/film formation. Yang et al. [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e] developed a CFD framework to represent phase expansion and wall impact and compared static/dynamic simulations on planar and curved surfaces with laboratory measurements. The study is useful for understanding how geometry shapes film formation; however, it does not quantify predictive accuracy in a way that supports day-to-day process decisions. Li et al. [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e] reported closer quantitative alignment between simulation and measurements for airless panels (e.g., comparable mean thickness and high transfer efficiency under their test conditions), yet comprehensive statistical validation across diverse operating conditions and simplified control-oriented rules were not provided. Wu et al. [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e] further analyzed curvature and path-planning effects for complex surfaces; while the visual correspondence is informative, accuracy metrics such as MAE/RMSE are generally not emphasized, and computational costs remain non-trivial.\u003c/p\u003e \u003cp\u003eMore recently, hybrid and control-oriented approaches have begun to appear. Shi et al. [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e] proposed a variable-flow model with speed planning that explicitly accounts for acceleration and deceleration during gun motion, thereby reducing local over- and under-coating compared with constant-speed/flow baselines. This highlights the importance of motion dynamics in thickness uniformity.\u003c/p\u003e \u003cp\u003ePrior studies share several limitations that constrain practical deployment. Factor spaces were defined narrowly, and motion dynamics and cross-factor interactions were often omitted. Experimental setups were mostly simplified, leaving robustness to variability in the real operation environment uncertain. They reported how well the model fit the experimental data, not predictive statistics with quantified uncertainty, and the models appeared to work within the tested setup and provided limited evidence of transfer to other operating conditions. Moreover, they ended with mechanism description or specific case studies and did not turn the results into simple rules that can be used for real operation.\u003c/p\u003e \u003cp\u003eThis study addresses these gaps by designing a five-factor central composite design (CCD) that captures pressure, thinning ratio, tip, distance, and speed under realistic production conditions. ANOVA is applied to select statistically significant main, interaction, and quadratic effects, producing concise and easy to interpret equations. Predictive validity was quantified using Pred. R\u003csup\u003e2\u003c/sup\u003e and related statistics, and actionable indicators for the shipyard environment were derived to enable direct parameter tuning without heavy computation. This framework keeps regression simple and low cost, yet delivers enough accuracy for real-time process guidance in complex shipyard environments. Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e summarizes the differences between this study and other references.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eSummary of the differences between this study and other references\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStudy cases\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTechnology and\u003c/p\u003e \u003cp\u003eapproach\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eKey focus and\u003c/p\u003e \u003cp\u003econtributions\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eLimitations\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRef. [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePlasma spray, regression (CCD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eLinked gas flow, powder feed, arc current, distance to coating outcomes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMulti-response focus reduced DFT-centric optimization; plasma-specific variability\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRef. [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHVLP, regression (DoE)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eEffect of nozzle size, spray time, air pressure on DFT\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eExcluded dynamic factors\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRef. [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRotational spray, regression (DoE)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eEffects of nozzle rotation, spindle speed, spray time\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eAssumes a fixed gun; limited transfer to handheld, dynamic field conditions\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRef. [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePowder coating, regression\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOptimization with current, voltage, substrate type\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eNarrow variable set; excluded spray dynamics\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRef. [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAirless spray, CFD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSimulated film formation, compared with experiments\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eComputationally intensive; generalized accuracy metrics (e.g., MAE/RMSE) limited\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRef. [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRotary airless spray, CFD-DEM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDroplet collision and accumulation modeling\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eNo statistical validation; no simplified rules\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRef. [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAirless spray on spherical surface, CFD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eModeled curvature and trajectory effects\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eDid not quantify predictive accuracy for practice\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRef. [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eVariable-velocity airless spray\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eModeled velocity-profile effects; improved uniformity vs constant baselines\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eImproved uniformity but no closed-form predictive equations\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAirless spray, regression (CCD)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ePredictive equations for FLOW, DFW, DFT, DFR in shipyard coatings\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e"},{"header":"3 Methodology","content":"\u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eIn this study, a DoE approach was employed to develop an airless spray model using a statistical methodology. The regression model for predicting FLOW, DFW, DFT, and DFR was derived from a dataset collected through an actual spray test with marine coating material commonly used in shipyards. The research was conducted in five stages: (1) defining the parameters and responses, (2) experimental design with DoE, (3) actual spray test with marine coating material, (4) data processing, and (5) model development. A schematic overview of the study is provided in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Definition of parameters and responses\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eIn this study, five process parameters, thinning ratio (THR), pump inlet gauge pressure (PRS), tip nozzle size (TIP), stand-off distance (DIST), and gun travel speed (SPD), were defined as system-controllable settings because they can be directly adjusted in the shipyard to regulate the operating state of airless spray coating operations. These controllable settings represent practical decision variables used by operators and supervisors to control spray property, represented by FLOW, and coating properties, represented by DFW, DFT, and DFR.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Experimental design\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eA five-factor CCD was generated in MINITAB 17 with a two-level full-factorial core and axial augmentation (α\u0026thinsp;=\u0026thinsp;2). The design comprised 80 cube points, 40 axial points, and 35 center points, resulting in a total of 155 runs in a single block. The continuous factors were studied at five coded levels (\u0026minus;\u0026thinsp;2, \u0026minus;\u0026thinsp;1, 0, +\u0026thinsp;1, +2), corresponding to the uncoded settings in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, whereas TIP was treated as a categorical factor with five levels. CCD enhances predictive reliability and enables robust estimation of curvature, thereby offering superior flexibility and precision compared to other second-order designs [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eLevels of the five factors in the CCD\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"8\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eParameters\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eType\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eSubtype\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"5\" nameend=\"c8\" namest=\"c4\"\u003e \u003cp\u003eCoded level\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026minus;2\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u0026minus;1\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003e+\u0026thinsp;1\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003e+\u0026thinsp;2\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTHR (vol.%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNumeric\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eContinuous\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e20\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePRS (bar)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNumeric\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eContinuous\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e3.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e4.50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e5.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e6.00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTIP (no.)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCategorical\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOrdinal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e519\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e523\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e527\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e531\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e535\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDIST (mm)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNumeric\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eContinuous\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e300\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e350\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e400\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e450\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e500\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSPD (mm/s)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNumeric\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eContinuous\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e400\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e550\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e700\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e850\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e1,000\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eAmong the studied factors, only the nozzle designation (TIP) is categorical while others are continuous variables. The TIP number is a manufacturer\u0026rsquo;s code that reflects the size of the elliptical orifice, which increases as the TIP number increases. Representative nozzle geometries for the TIP used in this work are illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e3.3 Actual spray test\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eIn the spray tests, an epoxy primer commonly applied in shipyards was used and diluted with a compatible thinner to achieve the target spray viscosity. The experimental setup consisted of an airless pump, prepared substrate specimens, in-line sensors, and a robot-mounted automatic spray gun. Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e summarizes the key characteristics of the main components of the experimental system.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eCharacteristics of the main components of the experimental system\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eComponents\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eParameter\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSpecification\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"4\" rowspan=\"5\"\u003e \u003cp\u003eCoating material\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePaint product name\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eJotacote universal N10\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eThinner product name\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eJotun thinner no.17\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eManufacturer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eJotun\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSolid volume ratio\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e72%\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eThinning ratio\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4\u0026ndash;20 vol.%\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eSubstrate specimen\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMaterial\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCarbon steel\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDimension\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e300 mm (width) \u0026times; 800 mm (height) \u0026times; 2 mm (thickness)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"3\" rowspan=\"4\"\u003e \u003cp\u003eAirless pump system\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePump inlet gauge pressure\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.0\u0026ndash;6.0 bar\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePressure ratio\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e72:1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMax pressure\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e432 bar\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHose length\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e55 m\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eSensors\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFlow sensor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eRecord real-time FLOW at 5 Hz\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTemperature sensor\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eRecord real-time temperature at 5 Hz\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eSpray robot\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCartesian coordinate robot\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eHorizontal traverse at constant speed\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGun\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAutomatic airless spray\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eFor the spray experiments, the setup included an airless spray pump, a flow sensor, a temperature sensor, a Cartesian coordinate robot with an automatic spray gun, and an integrated system of hoses delivering the coating material to the spray nozzle (refer to Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). The robot was used to ensure repeatable stand-off distance and travel speed across runs. A flow sensor placed along the hose continuously monitored the flow rate of the coating material. During the experiments, the specimens were positioned at a consistent distance from the spray nozzle. The auto spray gun, mounted on the Cartesian robot, was programmed to move horizontally at a uniform speed while cyclically activating the spray to ensure even application. Prior to spraying, all test specimens were mechanically prepared to eliminate contaminants and surface irregularities.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e3.4 Data processing\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThe FLOW, representing the spray property, was assessed by the flow sensor located at the middle of the hoses. During each spray trial, the sensor recorded real-time flow rate data at a rate of 5 Hz as the spray gun maintained a constant travel speed, applied paint to the specimen, and subsequently stopped spraying. As demonstrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, the measured FLOW values exhibited a periodic pulsation pattern, which was attributed to the piston\u0026rsquo;s reciprocating motion within the airless pump. To ensure accuracy in average values, the FLOW data was averaged specifically in \u0026ldquo;the steady-state interval\u0026rdquo;, where the readings stabilized while the spray gun passed over the specimen and built up the coating film.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThe coating properties of DFW, DFT, and DFR were also individually measured. The DFW was determined by measuring the fan-pattern width (i.e., the major-axis length of the elliptical fan) using a length gauge. The upper and lower limits were visually identified at the onset boundaries of continuous film formation. DFT was measured on each coated specimen using a dry-film thickness gauge. DFT can vary widely across a surface because there is a variation in particle velocities within the fan pattern during spraying onto specimen surface [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. Therefore, relying on single-point measurements may not yield an accurate representation of actual coating thickness [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. Thus, it is recommended to take multiple measurements and calculate their arithmetic mean to estimate the average DFT in a given area [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. In this study, fifty measurements of DFT were taken across the specimen surface and averaged. DFR was also measured at fifty locations using a roughness gauge with a 15 mm sampling length, and the mean value was calculated. Among the available roughness parameters, Rz was selected because it characterizes surface roughness by averaging the heights of the five highest peaks and the depths of the five deepest valleys over the sampling length, as defined by DIN 4768 [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eIn Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, the parameter y\u003csub\u003ei\u003c/sub\u003e represents the distance between the global maximum and minimum heights measured in five consecutive sampling sections over the total sampling length. Consequently, the Rz parameter is calculated as the mean of these five y\u003csub\u003ei\u003c/sub\u003e measurements [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe measurement points selected for evaluating the three coating property parameters of DFW, DFT, and DFR across the specimen surface are illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e3.5 Model development\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eUsing MINITAB 17 software, regression models in mathematical form were developed based on the dataset obtained from 155 experimental runs designed through CCD. Stepwise regression analysis was conducted at a significance level of 0.05 to eliminate statistically insignificant terms and to refine the overall model structure. The selection of significant predictors ensured model parsimony without compromising predictive performance. The final regression equations were validated using key statistical indicators, such as R-squared (R\u003csup\u003e2\u003c/sup\u003e), adjusted R-squared (Adj. R\u003csup\u003e2\u003c/sup\u003e), and predicted R-squared (Pred. R\u003csup\u003e2\u003c/sup\u003e), which reflect the goodness-of-fit and the reliability of the model.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"4 Results and discussion","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Experimental results\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eAll experimental trials were conducted in the randomized run order generated by MINITAB 17. For brevity, Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e lists the first and last five runs.\u003c/p\u003e \u003cp\u003eAnalysis of variance (ANOVA) was performed to investigate how the experimental parameters affected the four response variables. A significance threshold of 0.05 was used to guide stepwise model selection, whereby statistically insignificant terms were systematically excluded. In this procedure, linear, quadratic, and two-way interaction effects among the five parameters were initially considered, and only significant predictors were retained to achieve a concise model without reducing predictive accuracy.\u003c/p\u003e \u003cp\u003eWithin the ANOVA framework, the degree of freedom (DF) represents the number of independent components available to estimate variability. The adjusted sum of squares (Adj SS) quantifies each factor\u0026rsquo;s contribution after controlling for other variables, while dividing this by the DF yields the adjusted mean square (Adj MS). The F-statistic, calculated as the ratio of Adj MS for a given factor to the mean square of the residual error, provides a test of significance. The p-value, in turn, indicates the probability of observing such an effect under the null hypothesis, with values below 0.05 confirming statistical significance. The \u0026ldquo;Error\u0026rdquo; term reflects residual variation unexplained by the model, whereas \u0026ldquo;Total\u0026rdquo; denotes the overall variability of the dataset. Taken together, these ANOVA components establish the evidential basis for term retention and provide the foundation for subsequent model evaluation. The residual error can be decomposed into lack-of-fit and pure error. Pure error reflects repeatability estimated from replicated design points, whereas lack-of-fit quantifies systematic deviation between the model and the mean response. Because the present design includes multiple center/replicated points, the lack-of-fit test can become sensitive. Therefore, model adequacy was additionally assessed using Pred. R\u0026sup2; and residual diagnostics. To ensure field applicability, the final predictive equations were expressed in uncoded (real-world) units.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eCCD runs and measured responses\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"10\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eTest runs\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"5\" nameend=\"c6\" namest=\"c2\"\u003e \u003cp\u003eParameters\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c10\" namest=\"c7\"\u003e \u003cp\u003eResponses\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTHR\u003c/p\u003e \u003cp\u003e(vol.%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003ePRS\u003c/p\u003e \u003cp\u003e(bar)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eTIP\u003c/p\u003e \u003cp\u003e(no.)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eDIST\u003c/p\u003e \u003cp\u003e(mm)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eSPD\u003c/p\u003e \u003cp\u003e(mm/s)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eFLOW\u003c/p\u003e \u003cp\u003e(L/min)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eDFW\u003c/p\u003e \u003cp\u003e(mm)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eDFT\u003c/p\u003e \u003cp\u003e(\u0026micro;m)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003eDFR\u003c/p\u003e \u003cp\u003e(\u0026micro;m)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4.50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e527\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e400\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e700\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e2.05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e390\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e89.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e17.7\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4.50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e523\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e400\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1,000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1.59\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e513\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e48.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e30.6\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e523\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e350\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e550\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1.36\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e409\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e81.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e22.6\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e527\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e450\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e550\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1.49\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e420\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e83.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e28.9\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e535\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e450\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e850\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e2.60\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e480\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e77.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e25.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e⁝\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e⁝\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e⁝\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e⁝\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e⁝\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e⁝\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e⁝\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e⁝\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e⁝\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e⁝\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e151\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e523\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e350\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e550\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1.57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e445\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e82.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e20.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e152\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e527\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e350\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e850\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e2.04\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e371\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e90.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e22.2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e153\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e535\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e350\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e550\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1.72\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e301\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e145.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e23.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e154\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e531\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e350\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e850\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e3.19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e400\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e106.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e22.2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e155\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e535\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e350\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e850\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e2.44\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e318\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e120.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e24.7\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section3\"\u003e \u003ch2\u003e4.1.1 FLOW\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eAccording to ANOVA summarized in Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, the regression model for FLOW was statistically significant. The main effects of TIP, PRS, and THR were all highly significant, and two interaction terms of THR \u0026times; TIP and PRS \u0026times; TIP also showed significant contributions.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab5\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eANOVA of FLOW response\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSource\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDF\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAdj SS\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eAdj MS\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cem\u003eF\u003c/em\u003e-value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cem\u003ep\u003c/em\u003e-value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eModel\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e62.4179\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e3.90112\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e553.08\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLinear\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e57.4272\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e9.57121\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1356.95\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTHR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e9.5688\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e9.56884\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1356.62\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePRS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e9.1246\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e9.12457\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1293.63\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTIP\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e38.7338\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e9.68346\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1372.86\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSquare\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.0291\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.02909\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e4.12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.044\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTHR \u0026times; THR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.0291\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.02909\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e4.12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.044\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2-Way interactions\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4.9616\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.55128\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e78.16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTHR \u0026times; PRS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.1253\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.12529\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e17.76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTHR \u0026times; TIP\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3.3621\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.84051\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e119.16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePRS \u0026times; TIP\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1.4742\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.36855\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e52.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eError\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e138\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.9734\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.00705\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLack-of-Fit\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e108\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.8407\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.00778\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.038\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePure Error\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.1327\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.00442\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e154\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e63.3913\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eAmong these, TIP was found to have the most dominant influence on FLOW, as it directly affects the cross-sectional area through which the coating material passes. Among the interaction terms, THR\u0026times;TIP exhibited the largest effect, while PRS\u0026times;TIP and THR\u0026times;PRS also remained statistically significant. The uncoded regression equations for FLOW, corresponding to each TIP nozzle size, are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab6\" class=\"InternalRef\"\u003e6\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab6\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 6\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eUncoded regression equations for FLOW by TIP\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTIP\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eUncoded regression equation for FLOW (L/min)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e519\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.880\u0026thinsp;\u0026minus;\u0026thinsp;0.0192\u0026middot;THR\u0026thinsp;+\u0026thinsp;0.0060\u0026middot;PRS\u0026thinsp;\u0026minus;\u0026thinsp;0.000879\u0026middot;THR\u0026sup2; + 0.01319\u0026middot;THR\u0026middot;PRS\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e523\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.751\u0026thinsp;\u0026minus;\u0026thinsp;0.0035\u0026middot;THR\u0026thinsp;+\u0026thinsp;0.0746\u0026middot;PRS\u0026thinsp;\u0026minus;\u0026thinsp;0.000879\u0026middot;THR\u0026sup2; + 0.01319\u0026middot;THR\u0026middot;PRS\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e527\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.009\u0026thinsp;+\u0026thinsp;0.0276\u0026middot;THR\u0026thinsp;+\u0026thinsp;0.2433\u0026middot;PRS\u0026thinsp;\u0026minus;\u0026thinsp;0.000879\u0026middot;THR\u0026sup2; + 0.01319\u0026middot;THR\u0026middot;PRS\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e531\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.469\u0026thinsp;+\u0026thinsp;0.0619\u0026middot;THR\u0026thinsp;+\u0026thinsp;0.3278\u0026middot;PRS\u0026thinsp;\u0026minus;\u0026thinsp;0.000879\u0026middot;THR\u0026sup2; + 0.01319\u0026middot;THR\u0026middot;PRS\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e535\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;0.826\u0026thinsp;+\u0026thinsp;0.0949\u0026middot;THR\u0026thinsp;+\u0026thinsp;0.3951\u0026middot;PRS\u0026thinsp;\u0026minus;\u0026thinsp;0.000879\u0026middot;THR\u0026sup2; + 0.01319\u0026middot;THR\u0026middot;PRS\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThe model exhibited the highest predictive performance among all four responses, with a Pred. R\u0026sup2; of 0.9805, along with R\u0026sup2; = 0.9846 and Adj. R\u0026sup2; = 0.9829. These values indicate that the regression model is highly reliable in estimating FLOW under a wide range of operating conditions. In practical terms, FLOW increased with higher values of PRS and THR, as well as with the use of larger nozzle sizes. These relationships are intuitive, as greater pressure and lower viscosity facilitate increased spray volume of coating material.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section3\"\u003e \u003ch2\u003e4.1.2 DFW\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThe regression model for DFW also demonstrated strong statistical significance, with ANOVA results shown in Table\u0026nbsp;\u003cspan refid=\"Tab7\" class=\"InternalRef\"\u003e7\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab7\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 7\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eANOVA of DFW response\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSource\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDF\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAdj SS\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eAdj MS\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cem\u003eF\u003c/em\u003e-value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cem\u003ep\u003c/em\u003e-value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eModel\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e683,058\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e42,691\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e142.45\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLinear\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e651,849\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e93,121\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e310.72\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTHR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e48,088\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e48,088\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e160.46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePRS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1,397\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1,397\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e4.66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.033\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDIST\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e271,758\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e271,758\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e906.78\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTIP\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e330,606\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e82,652\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e275.78\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2-Way interactions\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e31,209\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e3,468\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e11.57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTHR \u0026times; PRS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1,373\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1,373\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e4.58\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.034\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTHR \u0026times; TIP\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e24,260\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e6,065\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e20.24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDIST \u0026times; TIP\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5,576\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1,394\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e4.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eError\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e138\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e41,358\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e300\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLack-of-Fit\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e108\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e40,632\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e376\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e15.54\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePure Error\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e726\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e154\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e724,416\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThe most influential parameter was DIST, followed by TIP, THR, and PRS. Additionally, the interaction between THR and TIP was statistically significant. The strong effect of stand-off distance on film width is consistent with the physical behavior of spray dispersion. The regression equations in uncoded units for each TIP nozzle are presented in Table\u0026nbsp;\u003cspan refid=\"Tab8\" class=\"InternalRef\"\u003e8\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab8\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 8\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eUncoded regression equations for DFW by TIP\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTIP\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eUncoded regression equation for DFW (mm)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e519\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;47.8\u0026thinsp;+\u0026thinsp;6.86\u0026middot;THR\u0026thinsp;+\u0026thinsp;21.12\u0026middot;PRS\u0026thinsp;+\u0026thinsp;1.0812\u0026middot;DIST\u0026thinsp;\u0026minus;\u0026thinsp;1.381\u0026middot;THR\u0026middot;PRS\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e523\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;112.1\u0026thinsp;+\u0026thinsp;10.63\u0026middot;THR\u0026thinsp;+\u0026thinsp;21.12\u0026middot;PRS\u0026thinsp;+\u0026thinsp;1.1507\u0026middot;DIST\u0026thinsp;\u0026minus;\u0026thinsp;1.381\u0026middot;THR\u0026middot;PRS\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e527\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;51.4\u0026thinsp;+\u0026thinsp;8.79\u0026middot;THR\u0026thinsp;+\u0026thinsp;21.12\u0026middot;PRS\u0026thinsp;+\u0026thinsp;0.8212\u0026middot;DIST\u0026thinsp;\u0026minus;\u0026thinsp;1.381\u0026middot;THR\u0026middot;PRS\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e531\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;118.7\u0026thinsp;+\u0026thinsp;12.60\u0026middot;THR\u0026thinsp;+\u0026thinsp;21.12\u0026middot;PRS\u0026thinsp;+\u0026thinsp;0.8565\u0026middot;DIST\u0026thinsp;\u0026minus;\u0026thinsp;1.381\u0026middot;THR\u0026middot;PRS\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e535\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;171.6\u0026thinsp;+\u0026thinsp;17.22\u0026middot;THR\u0026thinsp;+\u0026thinsp;21.12\u0026middot;PRS\u0026thinsp;+\u0026thinsp;0.8493\u0026middot;DIST\u0026thinsp;\u0026minus;\u0026thinsp;1.381\u0026middot;THR\u0026middot;PRS\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eModel evaluation metrics indicated a high level of fit, with R\u0026sup2; = 0.9429, Adj. R\u0026sup2; = 0.9363, and Pred. R\u0026sup2; = 0.9217. These results confirm that the regression model accurately describes the relationship between operating parameters and DFW. As expected, increased spray distance led to wider fan patterns and thus increased DFW. Conversely, larger nozzle sizes (higher TIP codes) tended to reduce DFW, which corresponds to narrower spray angles for larger nozzles. According to Lefebvre [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e], increasing the nozzle orifice diameter leads to a thicker annular liquid sheet at the exit, which generates larger ligaments and droplets via breakup mechanisms. However, such thick sheets are less likely to spread laterally, as their inertia suppresses sheet widening. As a result, the spray angle narrows, and the overall fan width is reduced despite the increase in flow rate. This theoretical explanation is consistent with our empirical observation that higher TIP numbers yield narrower dry film width.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section3\"\u003e \u003ch2\u003e4.1.3 DFT\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eAmong the four responses, the regression model for DFT is of greatest practical importance, particularly because of its relevance to regulatory compliance under IMO PSPC. As summarized in Table\u0026nbsp;\u003cspan refid=\"Tab9\" class=\"InternalRef\"\u003e9\u003c/span\u003e, all five main effects of THR, PRS, TIP, DIST, and SPD were statistically significant, with SPD and TIP exerting the strongest main effects on DFT. In addition, five two-way interaction terms also showed significant influence, with the SPD \u0026times; TIP interaction being the most pronounced, indicating complex interdependencies between the operating parameters.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab9\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 9\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eANOVA of DFT response\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSource\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDF\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAdj SS\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eAdj MS\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cem\u003eF\u003c/em\u003e-value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cem\u003ep\u003c/em\u003e-value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eModel\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e191,100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e8,308.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e221.17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLinear\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e178,206\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e22,275.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e592.96\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTHR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1,037\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1,037.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e27.61\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePRS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e13,951\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e13,951.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e371.36\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDIST\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e11,067\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e11,067.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e294.60\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSPD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e44,415\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e44,415.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1,182.29\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTIP\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e107,736\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e26,933.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e716.95\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSquare\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2,283\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2,282.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e60.76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSPD \u0026times; SPD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2,283\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2,282.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e60.76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2-Way interactions\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e10,611\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e757.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e20.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePRS \u0026times; SPD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e453\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e452.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e12.05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePRS \u0026times; TIP\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3,223\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e805.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e21.45\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDIST \u0026times; SPD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e526\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e525.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e14.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDIST \u0026times; TIP\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1,295\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e323.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e8.62\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSPD \u0026times; TIP\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5,115\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1,278.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e34.04\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eError\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e131\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4,921\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e37.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLack-of-Fit\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e101\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4,872\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e48.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e29.36\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePure Error\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e49\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e154\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e196,022\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThe TIP-specific uncoded regression equations are listed in Table\u0026nbsp;\u003cspan refid=\"Tab10\" class=\"InternalRef\"\u003e10\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab10\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 10\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eUncoded regression equations for DFT by TIP\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTIP\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eUncoded regression equation for DFT (\u0026micro;m)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e519\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e221.9\u0026thinsp;+\u0026thinsp;0.735\u0026middot;THR\u0026thinsp;+\u0026thinsp;20.22\u0026middot;PRS\u0026thinsp;\u0026minus;\u0026thinsp;0.3434\u0026middot;DIST\u0026thinsp;\u0026minus;\u0026thinsp;0.3542\u0026middot;SPD\u0026thinsp;+\u0026thinsp;0.000175\u0026middot;SPD\u0026sup2; \u0026minus; 0.02115\u0026middot;PRS\u0026middot;SPD\u0026thinsp;+\u0026thinsp;0.000342\u0026middot;DIST\u0026middot;SPD\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e523\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e249.6\u0026thinsp;+\u0026thinsp;0.735\u0026middot;THR\u0026thinsp;+\u0026thinsp;22.90\u0026middot;PRS\u0026thinsp;\u0026minus;\u0026thinsp;0.3679\u0026middot;DIST\u0026thinsp;\u0026minus;\u0026thinsp;0.3772\u0026middot;SPD\u0026thinsp;+\u0026thinsp;0.000175\u0026middot;SPD\u0026sup2; \u0026minus; 0.02115\u0026middot;PRS\u0026middot;SPD\u0026thinsp;+\u0026thinsp;0.000342\u0026middot;DIST\u0026middot;SPD\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e527\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e309.8\u0026thinsp;+\u0026thinsp;0.735\u0026middot;THR\u0026thinsp;+\u0026thinsp;29.77\u0026middot;PRS\u0026thinsp;\u0026minus;\u0026thinsp;0.4464\u0026middot;DIST\u0026thinsp;\u0026minus;\u0026thinsp;0.4243\u0026middot;SPD\u0026thinsp;+\u0026thinsp;0.000175\u0026middot;SPD\u0026sup2; \u0026minus; 0.02115\u0026middot;PRS\u0026middot;SPD\u0026thinsp;+\u0026thinsp;0.000342\u0026middot;DIST\u0026middot;SPD\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e531\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e338.9\u0026thinsp;+\u0026thinsp;0.735\u0026middot;THR\u0026thinsp;+\u0026thinsp;34.10\u0026middot;PRS\u0026thinsp;\u0026minus;\u0026thinsp;0.4878\u0026middot;DIST\u0026thinsp;\u0026minus;\u0026thinsp;0.4473\u0026middot;SPD\u0026thinsp;+\u0026thinsp;0.000175\u0026middot;SPD\u0026sup2; \u0026minus; 0.02115\u0026middot;PRS\u0026middot;SPD\u0026thinsp;+\u0026thinsp;0.000342\u0026middot;DIST\u0026middot;SPD\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e535\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e358.6\u0026thinsp;+\u0026thinsp;0.735\u0026middot;THR\u0026thinsp;+\u0026thinsp;38.90\u0026middot;PRS\u0026thinsp;\u0026minus;\u0026thinsp;0.5113\u0026middot;DIST\u0026thinsp;\u0026minus;\u0026thinsp;0.4717\u0026middot;SPD\u0026thinsp;+\u0026thinsp;0.000175\u0026middot;SPD\u0026sup2; \u0026minus; 0.02115\u0026middot;PRS\u0026middot;SPD\u0026thinsp;+\u0026thinsp;0.000342\u0026middot;DIST\u0026middot;SPD\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eAlthough the Pred. R\u0026sup2; for DFT was 0.9620, slightly lower than that of FLOW, the model still exhibited excellent predictive capacity (R\u0026sup2; = 0.9749, Adj. R\u0026sup2; = 0.9705), making it a robust tool for field implementation. From a practical perspective, DFT increased with higher values of PRS and larger nozzle sizes, while it decreased with greater DIST and SPD. The magnitude of the effect from TIP and SPD was particularly large. These findings suggest that achieving the nominal dry film thickness (nDFT) requires coordinated adjustment of not only the material delivery (e.g., PRS, THR) but also the spray gun movement dynamics (e.g., SPD, DIST). Therefore, the DFT model serves as the most critical foundation for establishing field-oriented spray control strategies.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section3\"\u003e \u003ch2\u003e4.1.4 DFR\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eFor DFR, Table\u0026nbsp;\u003cspan refid=\"Tab11\" class=\"InternalRef\"\u003e11\u003c/span\u003e shows that THR and SPD are the dominant main effects, while PRS, DIST, and TIP make smaller yet still significant contributions. The response is further shaped by three significant two-way interactions, with THR \u0026times; PRS exerting the strongest influence, followed by THR \u0026times; DIST and THR \u0026times; SPD.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab11\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 11\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eANOVA of DFR response\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSource\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDF\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAdj SS\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eAdj MS\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cem\u003eF\u003c/em\u003e-value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cem\u003ep\u003c/em\u003e-value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eModel\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4042.15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e288.72\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e40.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLinear\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3143.05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e392.88\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e55.35\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTHR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1384.34\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1384.34\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e195.02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePRS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e474.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e474.82\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e66.89\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDIST\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e235.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e235.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e33.13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSPD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e685.31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e685.31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e96.55\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTIP\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e363.41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e90.85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e12.80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSquare\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e323.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e323.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e45.60\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTHR \u0026times; THR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e323.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e323.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e45.60\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2-Way interactions\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e575.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e115.08\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e16.21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTHR \u0026times; PRS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e198.06\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e198.06\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e27.90\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTHR \u0026times; DIST\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e159.14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e159.14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e22.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTHR \u0026times; SPD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e122.78\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e122.78\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e17.30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePRS \u0026times; DIST\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e60.66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e60.66\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e8.55\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.004\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePRS \u0026times; SPD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e34.78\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e34.78\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e4.90\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.028\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eError\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e140\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e993.76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e7.10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLack-of-Fit\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e110\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e986.11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e8.96\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e35.15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePure Error\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e7.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e154\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5035.91\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThe regression equations for DFR in uncoded units are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab12\" class=\"InternalRef\"\u003e12\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab12\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 12\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eUncoded regression equations for DFR by TIP\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTIP\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eUncoded regression equation for DFR (\u0026micro;m)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e519\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;47.1\u0026thinsp;\u0026minus;\u0026thinsp;1.168\u0026middot;THR\u0026thinsp;+\u0026thinsp;4.44\u0026middot;PRS\u0026thinsp;+\u0026thinsp;0.2171\u0026middot;DIST\u0026thinsp;+\u0026thinsp;0.0671\u0026middot;SPD\u0026thinsp;+\u0026thinsp;0.0927\u0026middot;THR\u0026sup2; + 0.5245\u0026middot;THR\u0026middot;PRS\u0026thinsp;\u0026minus;\u0026thinsp;0.00705\u0026middot;THR\u0026middot;DIST\u0026thinsp;\u0026minus;\u0026thinsp;0.002065\u0026middot;THR\u0026middot;SPD\u0026thinsp;\u0026minus;\u0026thinsp;0.02322\u0026middot;PRS\u0026middot;DIST\u0026thinsp;\u0026minus;\u0026thinsp;0.00586\u0026middot;PRS\u0026middot;SPD\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e523\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;45.9\u0026thinsp;\u0026minus;\u0026thinsp;1.168\u0026middot;THR\u0026thinsp;+\u0026thinsp;4.44\u0026middot;PRS\u0026thinsp;+\u0026thinsp;0.2171\u0026middot;DIST\u0026thinsp;+\u0026thinsp;0.0671\u0026middot;SPD\u0026thinsp;+\u0026thinsp;0.0927\u0026middot;THR\u0026sup2; + 0.5245\u0026middot;THR\u0026middot;PRS\u0026thinsp;\u0026minus;\u0026thinsp;0.00705\u0026middot;THR\u0026middot;DIST\u0026thinsp;\u0026minus;\u0026thinsp;0.002065\u0026middot;THR\u0026middot;SPD\u0026thinsp;\u0026minus;\u0026thinsp;0.02322\u0026middot;PRS\u0026middot;DIST\u0026thinsp;\u0026minus;\u0026thinsp;0.00586\u0026middot;PRS\u0026middot;SPD\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e527\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;49.8\u0026thinsp;\u0026minus;\u0026thinsp;1.168\u0026middot;THR\u0026thinsp;+\u0026thinsp;4.44\u0026middot;PRS\u0026thinsp;+\u0026thinsp;0.2171\u0026middot;DIST\u0026thinsp;+\u0026thinsp;0.0671\u0026middot;SPD\u0026thinsp;+\u0026thinsp;0.0927\u0026middot;THR\u0026sup2; + 0.5245\u0026middot;THR\u0026middot;PRS\u0026thinsp;\u0026minus;\u0026thinsp;0.00705\u0026middot;THR\u0026middot;DIST\u0026thinsp;\u0026minus;\u0026thinsp;0.002065\u0026middot;THR\u0026middot;SPD\u0026thinsp;\u0026minus;\u0026thinsp;0.02322\u0026middot;PRS\u0026middot;DIST\u0026thinsp;\u0026minus;\u0026thinsp;0.00586\u0026middot;PRS\u0026middot;SPD\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e531\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;48.4\u0026thinsp;\u0026minus;\u0026thinsp;1.168\u0026middot;THR\u0026thinsp;+\u0026thinsp;4.44\u0026middot;PRS\u0026thinsp;+\u0026thinsp;0.2171\u0026middot;DIST\u0026thinsp;+\u0026thinsp;0.0671\u0026middot;SPD\u0026thinsp;+\u0026thinsp;0.0927\u0026middot;THR\u0026sup2; + 0.5245\u0026middot;THR\u0026middot;PRS\u0026thinsp;\u0026minus;\u0026thinsp;0.00705\u0026middot;THR\u0026middot;DIST\u0026thinsp;\u0026minus;\u0026thinsp;0.002065\u0026middot;THR\u0026middot;SPD\u0026thinsp;\u0026minus;\u0026thinsp;0.02322\u0026middot;PRS\u0026middot;DIST\u0026thinsp;\u0026minus;\u0026thinsp;0.00586\u0026middot;PRS\u0026middot;SPD\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e535\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026minus;45.8\u0026thinsp;\u0026minus;\u0026thinsp;1.168\u0026middot;THR\u0026thinsp;+\u0026thinsp;4.44\u0026middot;PRS\u0026thinsp;+\u0026thinsp;0.2171\u0026middot;DIST\u0026thinsp;+\u0026thinsp;0.0671\u0026middot;SPD\u0026thinsp;+\u0026thinsp;0.0927\u0026middot;THR\u0026sup2; + 0.5245\u0026middot;THR\u0026middot;PRS\u0026thinsp;\u0026minus;\u0026thinsp;0.00705\u0026middot;THR\u0026middot;DIST\u0026thinsp;\u0026minus;\u0026thinsp;0.002065\u0026middot;THR\u0026middot;SPD\u0026thinsp;\u0026minus;\u0026thinsp;0.02322\u0026middot;PRS\u0026middot;DIST\u0026thinsp;\u0026minus;\u0026thinsp;0.00586\u0026middot;PRS\u0026middot;SPD\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThe model for DFR exhibited acceptable performance, with R\u0026sup2; = 0.8027, Adj. R\u0026sup2; = 0.7829, and Pred. R\u0026sup2; = 0.7539. While lower than other response models, the performance was sufficient to provide meaningful process insights. Interpretation of the results showed that smoother surfaces (i.e., lower DFR values) were achieved under higher PRS and THR conditions, likely due to improved atomization and better surface leveling. In contrast, faster spray speeds and longer spray distances tended to increase surface roughness, likely due to incomplete film formation and reduced droplet overlap. This trend aligns with previous studies. Luangkularb et al. [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e] reported that increased spray distance reduces droplet momentum, leading to poor droplet coalescence and higher roughness. Similarly, Abe et al. [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e] observed that excessive spray velocity or stand-off distance impairs splat deformation and bonding, resulting in a more porous and irregular coating surface. These findings highlight the importance of controlling application dynamics in order to improve the surface quality of topcoats or anti-fouling coatings in marine applications.\u003c/p\u003e \u003cp\u003eResidual diagnostics (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e) indicated approximate normality and no major departures from model assumptions within the investigated region, with a slight increase in variance at higher fitted values.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e4.2 Discussion\u003c/h2\u003e \u003cdiv id=\"Sec16\" class=\"Section3\"\u003e \u003ch2\u003e4.2.1 Factor Effects on Spray and Coating Properties\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThe regression analysis quantified the influence of five operating parameters of THR, PRS, TIP, DIST, and SPD on four spray and coating properties: FLOW, DFW, DFT, and DFR. In addition to main effects, two-way interaction terms were selectively retained when statistically significant. The dominant contributing factors for each response are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab13\" class=\"InternalRef\"\u003e13\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab13\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 13\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eInfluential factors per response (main effects and key interactions)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eResponse\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDominant Factors (descending order of influence)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFLOW\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTIP\u0026thinsp;\u0026gt;\u0026thinsp;PRS\u0026thinsp;\u0026gt;\u0026thinsp;THR (interactions: THR\u0026times;TIP, PRS\u0026times;TIP)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDFW\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDIST\u0026thinsp;\u0026gt;\u0026thinsp;TIP\u0026thinsp;\u0026gt;\u0026thinsp;THR\u0026thinsp;\u0026gt;\u0026thinsp;PRS (interactions: THR\u0026times;TIP, DIST\u0026times;TIP)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDFT\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTIP\u0026thinsp;\u0026asymp;\u0026thinsp;SPD\u0026thinsp;\u0026gt;\u0026thinsp;PRS\u0026thinsp;\u0026gt;\u0026thinsp;DIST\u0026thinsp;\u0026gt;\u0026thinsp;THR (interactions: TIP\u0026times;SPD, DIST\u0026times;SPD, PRS\u0026times;TIP)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDFR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTHR\u0026thinsp;\u0026gt;\u0026thinsp;PRS\u0026thinsp;\u0026gt;\u0026thinsp;SPD\u0026thinsp;\u0026gt;\u0026thinsp;DIST (interactions: THR\u0026times;PRS, THR\u0026times;DIST, THR\u0026times;SPD)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section3\"\u003e \u003ch2\u003e4.2.2 Regression Model Performance\u003c/h2\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eUsing a Central Composite Design (CCD) with 155 experimental runs, regression models were developed for all four responses. Stepwise selection was employed to retain only statistically significant terms. The final models demonstrated strong predictive performance as summarized in Table\u0026nbsp;\u003cspan refid=\"Tab14\" class=\"InternalRef\"\u003e14\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab14\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 14\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eRegression model performance\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eResponses\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAdj. R\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003ePred. R\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFLOW\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.9846\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.9829\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.9805\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDFW\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.9429\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.9363\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.9217\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDFT\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.9749\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.9705\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.9620\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDFR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.8027\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.7829\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.7539\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThe FLOW, DFW, and DFT models all showed excellent fit and generalization capability, with Pred. R\u0026sup2; exceeding 92%. Although the DFR model had a lower Pred. R\u0026sup2;, the model still captures key trends and offers useful insights for managing surface quality in shipyard applications. To ensure practical usability, all regression equations were provided in uncoded units, and most were stratified by TIP. This enables direct application in shipyard environments without the need for transformation or post-processing.\u003c/p\u003e \u003cp\u003eFor a more comprehensive assessment of the proposed model\u0026rsquo;s predictive ability, its performance was benchmarked against regression-based DFT models reported in prior studies using Pred. R\u0026sup2; as a common reference metric. As summarized in Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e, the proposed model achieved Pred. R\u0026sup2; = 0.962 and compares favorably with the regression models of Datta et al. [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e], Luangkularb et al. [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e], Choikhrue et al. [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e], and Šolić et al. [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. These results support the robustness of the developed airless-spray model and indicate strong predictive reliability for industrial coating applications within the investigated operating window.\u003c/p\u003e \u003cp\u003eAlthough the lack-of-fit test indicated statistically detectable deviations from the second-order form, the equations were retained to preserve interpretability for field tuning. Within the investigated CCD region, predictive utility was supported by Pred. R\u0026sup2; and residual diagnostics (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e), and the equations should be used for prediction and parameter tuning within this region rather than extrapolated beyond it.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"5 Conclusions","content":"\u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThis study developed and validated regression models for four key outcomes of FLOW, DFW, DFT, and DFR for the airless spray coating process in shipbuilding using a five-factor Central Composite Design and stepwise term selection. The models were fitted to experimental data acquired under realistic shipyard operation conditions; only statistically significant main, interaction, and quadratic terms were retained, yielding concise yet accurate predictive equations.\u003c/p\u003e \u003cp\u003eThe factor-importance analysis showed that DFT is predominantly governed by TIP and SPD (followed by PRS, DIST, THR), DFW is driven mainly by DIST and TIP, FLOW is most sensitive to TIP and PRS (then THR), and DFR depends strongly on THR and PRS. Moreover, several interactions also play meaningful roles. The regression models exhibited strong predictive performance: Pred. R\u0026sup2; was 0.9805 for FLOW, 0.9217 for DFW, 0.9620 for DFT, and 0.7539 for DFR (with corresponding R\u0026sup2; and Adj. R\u0026sup2; values also high), indicating excellent fit and generalization for FLOW, DFW, and DFT, and practically useful trends for DFR.\u003c/p\u003e \u003cp\u003eTo facilitate direct field use, the final equations were expressed in uncoded (real-world) units and, where appropriate, stratified by nozzle TIP, enabling practitioners to set operating conditions without additional transformation or post-processing. Operators and supervisors can use the regression equations to tune process parameters, maintain target DFT, and achieve high film quality.\u003c/p\u003e \u003cp\u003eBenchmarking against selected prior regression-based DFT studies indicates that the present DFT model shows competitive predictive performance (Pred. R\u0026sup2; = 0.962) relative to predictive indicators reported in the literature. This result supports the robustness of the proposed approach and its practical utility for guiding coating operations within the investigated process window.\u003c/p\u003e \u003cp\u003eOverall, the work delivers an interpretable, high-accuracy, and deployment-ready predictive framework for airless spray coating in shipbuilding, offering clear parameter guidance to achieve target DFT while balancing fan width, flow, and surface quality.\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eEthics approval\u003c/h2\u003e\n\u003cp\u003eHerewith the confirmation: The paper was and is not submitted for publication elsewhere. This paper has not been published elsewhere in its entirety, in part, or in a modified version. The paper was not submitted for possible publication elsewhere.\u003c/p\u003e\n\u003ch2\u003eConsent for publication\u003c/h2\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003ch2\u003eConsent to participate\u003c/h2\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003ch2\u003eCompeting interests\u003c/h2\u003e\n\u003cp\u003eThe authors have no relevant financial or non-financial interests to disclose.\u003c/p\u003e\n\u003ch2\u003eFunding\u003c/h2\u003e\n\u003cp\u003eThe authors declare that no funds, grants, or other support were received for conducting this research.\u003c/p\u003e\n\u003ch2\u003eAuthor Contributions\u003c/h2\u003e\n\u003cp\u003eJinuk Kim: Conceptualization, Data curation, Formal analysis, Investigation, Software, Writing \u0026ndash; original draft. Kwangyeol Ryu: Methodology, Supervision, Validation, Writing \u0026ndash; review \u0026amp; editing. Yeonho Cho: Project administration, Resources. All authors have read and approved the final manuscript.\u003c/p\u003e\n\u003ch2\u003eData availability\u003c/h2\u003e\n\u003cp\u003eThe data that support the findings of this study are available from the corresponding author upon reasonable request, due to proprietary agreements with the collaborating shipyard.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eOriaifo E, Perera N, Guy A, Leung PS, Tan K (2014) A review of test protocols for assessing coating performance of water ballast tank coatings. In: The ICMNE 2014: XII International Conference on Marine and Naval Engineering, Istanbul, Turkey, 29\u0026ndash;30 September 2014. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5281/zenodo.1096381\u003c/span\u003e\u003cspan address=\"10.5281/zenodo.1096381\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIannarelli P, Beaumont D, Liu Y et al (2022) The degradation mechanism of a marine coating under service conditions of water ballast tank. Prog Org Coat 162:106588. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.porgcoat.2021.106588\u003c/span\u003e\u003cspan address=\"10.1016/j.porgcoat.2021.106588\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDe Baere K, Verstraelen H, Rigo P, Van Passel S, Lenaerts S, Potters G (2013) Reducing the cost of ballast tank corrosion: an economic modeling approach. Mar Struct 32:136\u0026ndash;152. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.marstruc.2012.10.009\u003c/span\u003e\u003cspan address=\"10.1016/j.marstruc.2012.10.009\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWillemen R, Luyckx D, Meskens R, Lenaerts S, De Baere K (2020) A study into the coating thickness of ship ballast tanks. Int J Mar Eng 162(A3):277\u0026ndash;287. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5750/ijme.v162iA3.1137\u003c/span\u003e\u003cspan address=\"10.5750/ijme.v162iA3.1137\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eInternational Maritime Organization (IMO) (2006) Resolution MSC.215(82): Performance Standard for Protective Coatings for Dedicated Seawater Ballast Tanks in All Types of Ships and Double-Side Skin Spaces of Bulk Carriers (adopted on 8 December 2006. IMO, London. effective 1 July 2008)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDatta S, Pratihar DK, Bandyopadhyay PP (2013) Modeling of plasma spray coating process using statistical regression analysis. Int J Adv Manuf Technol 65:967\u0026ndash;980. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00170-012-4232-y\u003c/span\u003e\u003cspan address=\"10.1007/s00170-012-4232-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLuangkularb S, Prombanpong S, Tangwarodomnukun V (2014) Material consumption and dry film thickness in spray coating process. Procedia CIRP 17:789\u0026ndash;794. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.procir.2014.02.046\u003c/span\u003e\u003cspan address=\"10.1016/j.procir.2014.02.046\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChoikhrue M, Prombanpong S, Sriyotha P (2016) An effect of coating parameters to dry film thickness in spray coating process. Key Engineering Materials 709:95\u0026ndash;98. https://doi.org\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e/10.4028/www.scientific.net/KEM.709.95\u003c/span\u003e\u003cspan address=\"http:///10.4028/www.scientific.net/KEM.709.95\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eŠolić T, Marić D, Peko I, Samardžić I (2025) Optimization of Coating Process Parameters by Analysis of Target Powder Thickness and Regression Modeling. Appl Sci 15(2):673. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/app15020673\u003c/span\u003e\u003cspan address=\"10.3390/app15020673\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang G, Wu Z, Chen Y, Chen S, Jiang J (2022) Modeling and characteristics of airless spray film formation. Coatings 12(7):949. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/coatings12070949\u003c/span\u003e\u003cspan address=\"10.3390/coatings12070949\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi X, Chen X, Hong N, Li Q, Xu Z, Sheng M, Wang R (2024) A CFD-DEM simulation of droplets in an airless spray coating process of a square duct. Coatings 14(3):282. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/coatings14030282\u003c/span\u003e\u003cspan address=\"10.3390/coatings14030282\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu Z, Wang C, Yang G, Chen S, Duan J, Chen Y (2024) Numerical investigation of film formation characteristics and mechanisms through airless spraying on spherical surfaces. Coatings 14(10):1299. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/coatings14101299\u003c/span\u003e\u003cspan address=\"10.3390/coatings14101299\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShi T, Xu J, Cui J, Tao L, Xu W, Wang Z, Ji J (2023) Variable velocity coating thickness distribution model for super-large planar robot spraying. Coatings 13(8):1434. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/coatings13081434\u003c/span\u003e\u003cspan address=\"10.3390/coatings13081434\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMyers RH, Montgomery DC, Anderson-Cook CM (2016) Response surface methodology: Process and product optimization using designed experiments, 4th edn. Wiley, Hoboken\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim SK, Jang SK, Kim DK, Jang JR (2011) Design and performance evaluation of airless nozzle for painting. Proceedings of the Korean Society for Precision Engineering Spring Conference (Part II), pp 1363\u0026ndash;1364\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eASTM D7091-22 (2022) Standard Practice for Nondestructive Measurement of Dry Film Thickness of Nonmagnetic Coatings Applied to Ferrous Metals and Nonmagnetic, Nonconductive Coatings Applied to Non-Ferrous Metals. ASTM International, West Conshohocken, PA, USA\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSSPC-PA 2 (2022) Procedure for Determining Conformance to Dry Coating Thickness Requirements. AMPP (Association for Materials Protection and Performance), Houston, TX, USA (2022)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDeutsches Institut f\u0026uuml;r Normung (DIN) (1990) DIN 4768: Determination of values of surface roughness parameters Ra, Rz, Rmax using electrical contact (stylus) instruments\u0026mdash;Concepts and measuring conditions\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAmaral MM, Raele MP, Caly JP, Samad RE, Vieira ND Jr, Freitas AZ (2009) Roughness measurement methodology according to DIN 4768 using optical coherence tomography (OCT). In: Modeling Aspects in Optical Metrology II. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1117/12.827748\u003c/span\u003e\u003cspan address=\"10.1117/12.827748\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLefebvre AH, McDonell VG (2017) Atomization and sprays, 2nd edn. CRC, Boca Raton\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAbe A, Goss JA, Zou M (2024) Exploring the impact of spray process parameters on graphite coatings: Morphology, thickness, and tribological properties. Coatings 14(6):714. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/coatings14060714\u003c/span\u003e\u003cspan address=\"10.3390/coatings14060714\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Shipbuilding, Protective coatings, Airless spray, Design of experiments, Regression model","lastPublishedDoi":"10.21203/rs.3.rs-8668336/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8668336/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eWater ballast tanks (WBTs) are exposed to severe cyclic corrosion in marine environments. The service life of a hull often depends on whether the protective coating in WBTs satisfies the dry film thickness (DFT) requirements specified in the IMO Performance Standard for Protective Coatings (PSPC). In shipyards, epoxy primers are applied mainly by manual airless spray, and thickness is difficult to control, leading to rework from insufficient thickness or material waste from excess thickness. This study develops interpretable regression models to predict and tune key airless spray outcomes under shipyard conditions. A five-factor Central Composite Design (155 runs) varied thinning ratio (THR), pump inlet gauge pressure (PRS), tip size (TIP), stand-off distance (DIST), and gun travel speed (SPD). Responses included coating flow rate (FLOW), dry film width (DFW), average DFT, and dry film roughness (DFR). Stepwise regression with analysis of variance (ANOVA) identified significant main, interaction, and quadratic effects and yielded uncoded predictive equations. Model assumptions were assessed using residual diagnostics, and predictive accuracy was evaluated using Pred. R\u0026sup2;. All five factors significantly influenced DFT, with SPD and TIP contributing most to thickness control. DFR was dominated by PRS, THR, SPD, and DIST, reflecting coupled effects of viscosity, atomization energy, and deposition distance. The final models agreed well with experiments across the design space. The equations enable forward prediction and parameter tuning for controlled multi-pass application to achieve target DFT while reducing the risk of insufficient or excess thickness and material loss.\u003c/p\u003e","manuscriptTitle":"Data-Driven Regression Model for Systematic Control of Airless Spray Coating in Shipbuilding Industry","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-17 15:02:09","doi":"10.21203/rs.3.rs-8668336/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"01812c57-8fe3-45e3-945b-66a75277bcf8","owner":[],"postedDate":"February 17th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-04-04T13:06:33+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-17 15:02:09","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8668336","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8668336","identity":"rs-8668336","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}