Field measurements were conducted using both mechanical and digital instruments in the air-blast freezer system. The collected data included temperature and pressure variables measured at several critical points within the system at specific time intervals throughout the operational periods (Hermes et al., 2021). Temperature monitoring was conducted using RTD PT100 sensors (Class A, ±0.15°C accuracy) with stainless steel probes (Omron E52 Series). The air velocity inside the freezing chamber was measured using a Krisbow digital anemometer (±3% accuracy, 0.01 m/s resolution) to quantify the airflow variations associated with the development of frosting and heat transfer performance. All measurements were recorded at fixed time intervals to capture the transient thermodynamic behavior during the early freezing stage. The primary objective of this measurement was to observe the system performance, identify the patterns of temperature and pressure fluctuations, and evaluate the overall system eficiency (Badri et al., 2021). The analysis and evaluation included:

1.Calculate the ABF air flow rate

Q = air flow rate capacity

A = cross-sectional area of air duct

V = air flow rate m/s

2.Efficiency of temperature distribution in ABF

air temperature entering the ABF room

outlet air temperature to the ABF room

3.Cooling capacity

Q= cooling capacity (kJ or kW)

m= frozen shrimp mass (Kg)

C = specific heat capacity of shrimp before freezing

∆t= initial shrimp temperature – shrimp temperature reaches freezing point (°C)

latent heat of freezing shrimp (kJ/Kg)

4.Frosting effect on the evaporator

Q= cooling capacity (kJ or kW)

m= frozen shrimp mass (Kg)

Cp= specific heat capacity of shrimp before freezing (kJ/Kg)

∆t= initial shrimp temperature – shrimp temperature reaches freezing point (°C)

5.Efficiency of the refrigeratoin system (n)

This eficiency is expressed as a percentage and calculated as the ratio between the actual coeficient of performance (COP) of the system and the Carnot COP, which represents the “theoretical” value derived from the condensation and evaporation temperatures (Srithar et al., 2025). The percentages of eficiency levels are presented in Table 4.

Shrimp samples were evaluated during the first 20 h of storage in an air blast freezer (ABF). All parameters were analyzed in triplicate at each interval. The pH was measured using a calibrated pH meter on homogenized tissue prepared at a 1:9 ratio, and texture (firmness) was assessed using a simple penetrometer applied to standardized muscle sections. Drip loss was determined gravimetrically by calculating the diference in the sample weight before and after controlled storage. Melanosis was visually evaluated using a simplified scoring scale. The water holding capacity (WHC) was quantified using a centrifugation-based method to determine the percentage of retained moisture. Ice crystal size was observed in thin frozen sections using microscopy. Total volatile base nitrogen (TVB-N) was measured using the Conway difusion method, followed by alkaline titration (Kim et al., 2020).

All thermodynamic performance and shrimp quality parameters were analyzed descriptively and statistically to evaluate the temporal changes during the transient air-blast freezing. The thermodynamic parameters included airflow velocity, air-side temperature diferential (ΔT), evaporator frosting mass, cooling capacity (Q), and the actual coeficient of performance (COP). The shrimp quality parameters included ice crystal size, water-holding capacity (WHC), drip loss, texture, pH, melanosis, and TVB-N. The data are presented in the Results and Discussion sections as tables and figures according to the characteristics of each variable. Descriptive analysis was used to summarize the direction, pattern, and magnitude of changes, whereas statistical analysis was applied, where appropriate, to assess diferences among freezing intervals and examine associations among measured parameters, particularly those related to frosting accumulation, airflow reduction, cooling capacity changes, COP variation, and shrimp quality responses.

Transient ABF Thermodynamic Behavior testing was carried out with three treatments at each time, as presented in Table 5.

Table 5 demonstrates a clear transient degradation of the ABF thermodynamic performance during the first 5 h of operation, characterized by a progressive decline in airflow , cooling capacity Q kW), and COP , concomitant with exponential frosting accumulation on the evaporator coil . The monotonic reduction in airflow indicates increasing aerodynamic resistance across the coil, which suppresses convective heat transfer and lowers the efective temperature gradient, thereby directly reducing the evaporator heat absorption and system energy eficiency (Boeng & Stahelin et al., 2025; de Sa Sarmiento et al., 2025). This coupled behavior confirms that early stage frost growth acts as a dominant transient thermal resistance layer, impairing both sensible and latent heat exchanges and accelerating COP deterioration (Jia et al., 2021; Kumar et al., 2024).

The nearly linear decline in Q and COP relative to the frosting mass suggests a strong thermodynamic sensitivity of the system performance to surface blockage during the pre-steady freezing phase. These trends are consistent with recent refrigeration system analyses reporting that frost layer thickening reduces air-side heat transfer coeficients and increases pressure drop, ultimately decreasing system capacity by 15–35% within short operating cycles (Hermes et al., 2021; P. Zheng et al., 2025; C. Ma et al., 2025). Furthermore, studies on air-blast and cold-room evaporators have confirmed that early transient frosting critically determines downstream product freezing kinetics owing to reduced convective intensity and altered thermal uniformity (C. Wang et al., 2022; S. Liu, Zhang, Li, et al., 2024). Therefore, the dataset robustly supports the interpretation that transient frosting accumulation is the primary mechanistic driver of early ABF performance decay, with direct implications for freezing rate stability and energy eficiency optimization in exportgrade shrimp-processing systems (Cezar et al., 2020; Yu et al., 2021; Ye et al., 2024).

Pearson correlation analysis, justified by Shapiro-Wilk normality testing 0 . 0 5 \end{document} ]]> for all variables), revealed an exceptionally strong thermodynamic coupling among the frosting mass, airflow, ΔT, evaporator cooling capacity (Q), and COP during the first 5 h of ABF operation. Frosting exhibited a near-perfect negative association with COP 9, p < 0.001) and airflow , whereas airflow was strongly correlated with Q (r = 0.994) and COP (r = 0.996), confirming a tightly synchronized degradation pathway (Alarcón-Gallén et al., 2025; Jiao et al., 2024). Linear regression further demonstrated that COP decline can be robustly predicted from frosting accumulation according to COP = 3.64 − 0.86(Frosting) with , indicating that nearly 98% of COP variance during early operation is explained by frost growth. Similarly, 0.973), substantiating that airflow obstruction induced by frost formation directly suppresses cooling capacity before steady-state conditions are reached (Klingebiel et al., 2025; Kumar et al., 2024).

These findings move beyond descriptive transient reporting by quantitatively resolving the mechanistic cascade responsible for the early stage performance decay. The process begins with frost deposition on the evaporator surface, which progressively restricts the airflow within the freezing chamber. This airflow restriction reduces the efective heat transfer gradient between the cooling air and the product, leading to the suppression of the cooling capacity of the system. The decline in cooling capacity subsequently results in a measurable degradation of the coeficient of performance (COP).

The results demonstrate that eficiency loss in industrial Air Blast Freezer (ABF) systems is not merely a time-dependent phenomenon but a statistically predictable process governed by the underlying mechanistic interactions occurring during the early stages of operation (Wang et al., 2024; Wu et al., 2024; Munk et al., 2026). The near-deterministic correlations observed in this study highlight a critical and previously under-quantified early operation vulnerability window, ofering a predictive framework that can be integrated with product quality kinetics to bridge machine performance dynamics and shrimp microstructural degradation (Food and Drug Administration, 2022; Xu et al., 2024; Zhang et al., 2025).

The changes in shrimp quality attributes during the first 6 h of ABF operation were evaluated to describe the early stage deterioration pattern associated with freezing performance (Table 8).

The time-resolved quality profile presented in Table 8 demonstrates a clear dynamic response of shrimp muscle properties during the first 6 h of ABF operation, characterized by an early rapid transition phase, followed by a progressive attenuation of change. At the onset (0-1 h), quality parameters exhibited moderate but consistent shifts, with ice crystal size increasing from 8 to 10 µm and WHC declining from 86% to 83%, accompanied by a slight rise in drip loss (1.80-2.00%) and a gradual reduction in texture firmness (11.0-10.4 N). This initial phase reflects the structural adjustment of muscle tissue during primary ice nucleation and crystal growth (Zhang et al., 2021; Zhang et al., 2020; Ahmad et al., 2025). Between 2-4 h, the rate of deterioration became more pronounced: ice crystals enlarged from 11 to 14 µm, WHC decreased further to 78%, and firmness dropped to 9.0 N, while drip loss increased to 2.40%. This period indicates intensified intracellular damage, likely associated with reduced freezing eficiency and increased thermal resistance within the system (Liu et al., 2024).

Notably, the melanosis score rose progressively (0.9-1.4), suggesting that oxidative and enzymatic reactions continued despite subzero conditions (Han & Gokoglu, 2022;(Kittiphattanabawon et al., 2024; Durage et al., 2025). After 4 h, the system entered a quasi-stagnation phase, where changes became marginal; the ice crystal size increased only slightly (15.0-15.2 µm), WHC decline slowed (77-76.8%), and drip loss plateaued (2.50- 2.52%). The stabilization of pH (7.11) and the minimal increment in TVB-N (18-18.2 mg N/100 g) further indicate that early structural degradation predominates over biochemical spoilage within this time window (Sun et al., 2023a; Alam et al., 2023c). Overall, the data reveal a distinct critical period within the first 2–4 h, in which ABF performance exerts the strongest influence on microstructural integrity, while beyond 5 h, the deterioration trend approaches a steady-state condition (Ji et al., 2021), highlighting the importance of controlling transient freezing eficiency during the early operational stage to preserve export-grade shrimp quality (Food and Drug Administration, 2022; Wei et al., 2024; Xu et al., 2025; Zheng et al., 2026).

Table 9 demonstrates a robust and sequential thermodynamic-microstructuralfunctional linkage during the first 0–6 h of ABF operation, confirming that system performance deterioration propagates mechanistically toward product quality loss. The regression results indicate that COP exerts a strong and significant negative influence on ice crystal size 0.97; p < 0.001), indicating that declining refrigeration eficiency directly promotes larger ice crystal formation (Minh, 2023). In turn, ice crystal size significantly reduced water-holding capacity (β = -1.17; R² = 0.96; p < 0.001), reflecting the structural disruption of muscle fibers and increased extracellular ice growth. Subsequently, WHC showed a strong inverse relationship with drip loss (β , confirming that reduced structural integrity translates into higher exudate release upon thawing (Sun et al., 2023a, 2023b). The total indirect pathway (COP → Ice Crystal → WHC → Drip Loss) exhibited high explanatory power , substantiating that COP does not merely represent an operational eficiency indicator but functions as a predictive upstream determinant of shrimp quality degradation through structural mediation (Klunklin et al., 2025; X. Zheng et al., 2026). Collectively, these findings provide quantitative evidence of a cascading cause-efect mechanism, positioning COP as an early thermodynamic control variable that governs microstructural stability and final functional quality during transient ABF performance.

The transient behavior of the ABF operation was analyzed in relation to ice crystal development to illustrate how changes in freezing performance influence the microstructural formation in shrimp products (Figure 1).Figure 1 illustrates the inverse dynamic relationship between the actual coeficient of performance (COP) of the Air Blast Freezer (ABF) and the evolution of ice crystal size during the first six hours of freezing. At the initial stage (0-1 h), the COP remains relatively high (3.5-3.2), corresponding to smaller ice crystal dimensions (8-10 µm), indicating eficient heat extraction and rapid surface freezing (Han & Gokoglu, 2022; Saini et al., 2021). As the operation progressed, a gradual decline in COP was observed, reaching approximately 2.3 at 6 h. This thermodynamic deterioration coincided with a steady increase in the ice crystal size to 15 µm. The data suggest that transient eficiency loss, likely associated with frosting accumulation and reduced heat transfer capacity, directly influences freezing kinetics, allowing slower intracellular water solidification and promoting larger crystal formation (S. Liu, Zhang, Li, et al., 2024; S et al., 2023). This pattern confirms that early stage ABF stability plays a decisive role in controlling the microstructural integrity of shrimp muscle during freezing, establishing a mechanistic link between machine performance and product quality (Tan et al., 2021; Yan et al., 2023).Figure 2 demonstrates a clear divergence between the water holding capacity (WHC) and drip loss throughout the first six hours of freezing. WHC showed a progressive decline from approximately 86% at the initial stage to below 77% at hour six, indicating a gradual impairment of the muscle’s ability to retain intracellular water (Li et al., 2024; Kamali et al., 2024). In parallel, drip loss increased consistently from approximately 1.8% to approximately 2.5%, reflecting intensified exudate release during thawing (Q. Sun, Zhang, Yang, Hou, et al., 2023; Yan et al., 2023). The near-linear and opposing trajectories of these two parameters indicate a strong inverse association between structural water retention and the physical moisture loss. The transition becomes more pronounced after approximately three hours, suggesting that cumulative microstructural damage, likely associated with enlarging ice crystals, compromises cellular integrity and accelerates extracellular fluid migration (Zhou et al., 2026; Klunklin et al., 2025). Collectively, the data confirm that transient freezing dynamics critically influence functional quality attributes, with a declining WHC serving as a precursor to measurable drip loss escalation.

Figure 3 presents the interaction between freezing time and system performance (COP) in terms of drip-loss development. During the high-COP phase (0-3 h), the drip loss increased gradually from approximately 1.8% to 2.3%, indicating controlled moisture migration under relatively stable thermodynamic conditions (Bao et al., 2023).

In contrast, once the system transitioned into the lower COP phase (≥3.5 h), the slope of the drip loss became steeper, rising to 2.5% by hour six. This divergence suggests that freezing time alone does not fully explain quality degradation; rather, the combined efects of prolonged exposure and reduced refrigeration eficiency amplify structural damage. This interaction pattern supports the hypothesis that early stage thermodynamic stability moderates moisture retention, whereas declining COP intensifies exudative losses (Skonieczny et al., 2026; Y. Zhao et al., 2025; K. Sun, Pan, Chen, Tao, et al., 2023; L. Liu et al., 2023; Phan et al., 2021). Thus, the time × COP interaction highlights a critical operational window in which maintaining a higher system eficiency can substantially mitigate quality deterioration.Figure 4 depicts the decomposition of the standardized efects linking COP to the drip loss through a mediation framework. The direct efect of COP on drip loss was negligible and slightly negative (-0.01), indicating that thermodynamic eficiency does not substantially influence moisture loss in the absence of structural intermediaries (Diao et al., 2021; Wei et al., 2024). In contrast, the indirect efect, quantified at approximately 0.265, dominated the relationship and closely approximated the total efect (0.255). This pattern confirms that the influence of COP on drip loss operates primarily through an intermediate variable, most plausibly ice crystal enlargement, and subsequent microstructural disruption (Z. Jin et al., 2024; O. Zheng et al., 2024). The near equivalence between the indirect and total efects suggests a strong mediation pathway, reinforcing the mechanistic chain whereby declining refrigeration eficiency alters freezing kinetics, modifies crystal morphology, and ultimately increases exudative losses. Collectively, the mediation analysis substantiates that product quality deterioration is not directly driven by energy performance alone, but by its downstream impact on the structural integrity.

Figure 5 shows the critical transition point in the drip loss development during the first six hours of freezing. Drip loss increased moderately from approximately 1.8% at initiation to approximately 2.3% by the third hour, reflecting relatively controlled structural stress under stable freezing conditions (Giannakourou & Dermesonlouoglou, 2024; Guo et al., 2022). However, beyond the \~3 h threshold, the trajectory approaches a steeper ascent toward ±2.5%, indicating accelerated exudative release (LAO et al., 2024; Wei et al., 2025). Therefore, the vertical demarcation at hour three therefore represents a pivotal operational window, after which cumulative thermodynamic ineficiencies and microstructural damage become more pronounced (Yan et al., 2023; Zhou et al., 2026; Kim et al., 2020). This inflection suggests that maintaining optimal system performance during the early freezing phase is essential to limit irreversible water migration.

Consequently, the data support the concept of a “critical freezing window” in which proactive control of heat transfer eficiency can significantly mitigate subsequent quality degradation.

The most decisive finding of this study is that early stage frost accumulation is the primary upstream driver of both thermodynamic degradation and shrimp microstructural deterioration during transient ABF operation. Frost growth during the first operational hours systematically reduced the airflow, suppressed the air-side temperature gradient, and diminished the cooling capacity and COP with near-deterministic statistical strength. These air-side reductions were synchronously associated with progressive ice crystal enlargement, declining water-holding capacity, and increasing drip loss, confirming that refrigeration eficiency functions as a quality-governing variable rather than merely an energy performance indicator (Wang Xu et al., 2024; da Silva Oliveira & Gonçalves, 2019; Sánchez-Vega et al., 2024).

Mechanistically, frost deposition increases the thermal resistance and pressure drop across the evaporator, thereby weakening the convective heat transfer intensity (P. Zheng et al., 2025). Reduced convective extraction plausibly slows surface heat removal from shrimp tissue, extending the product’s residence time within the critical crystallization temperature zone (approximately -1 to , where crystal growth kinetics are highly sensitive to the freezing rate. It is important to distinguish direct evidence from thermodynamic inference: airflow, COP, and structural quality parameters were empirically measured, whereas prolonged residence time in the critical zone was inferred from air-side degradation and crystal enlargement patterns (Alarcón-Gallén et al., 2025; Hermes et al., 2021). Although the product core temperature was not instrumentally monitored, the exceptionally strong inverse regression between the COP and ice crystal size supports the plausibility of this mechanistic pathway.

When compared critically with existing refrigeration and frozen seafood studies, the magnitude of the cooling capacity and COP reduction observed in this study aligns with previously reported frost-induced performance losses in industrial evaporators. However, most earlier investigations have treated frosting as an isolated energy eficiency issue or have evaluated shrimp microstructure independently of system thermodynamics. This study advances beyond descriptive alignment by statistically integrating machine performance and product structural response into a sequential mediation framework (COP → Ice Crystal → WHC → Drip Loss), thereby resolving the mechanistic cascade with high explanatory power. This integrative perspective reveals a distinct early operational vulnerability window (approximately 2-4 h), during which thermodynamic instability exerts the strongest influence on crystal morphology before quasi-stable conditions emerge.

From an operational and HACCP standpoint, airflow decline exceeding approximately one-third of the baseline, COP approaching 3.0, or frost mass nearing 1.0-1.2 kg may serve as practical early warning indicators of accelerated structural degradation. Nevertheless, while the regression pathways exhibit strong predictive capability, the mechanistic interpretation remains confined to measured air-side and structural variables; direct validation of residence time dynamics requires embedded core temperature monitoring and highresolution freezing rate analysis in future investigations.