Survival Losses Scale With Specific Mechanical Energy in Companion Animal Extrusion

Specific mechanical energy (SME) is defined as the net mechanical energy input per unit mass flow through the extruder and is computed from live torque, screw angular velocity, and mass flow rate according to SME (kJ/kg) = (τ_net × ω) / (1000 × ṁ), where τ_net is the net torque in N·m after subtracting empty-barrel frictional torque, ω is the screw angular velocity in rad/s, and ṁ is the mass flow rate in kg/s. In companion animal extrusion, production-scale co-rotating twin-screw extruders with L/D ratios ranging from 25:1 to 45:1 and screw diameters between 50 mm and 150 mm typically operate at SME values from 150 kJ/kg to 350 kJ/kg for standard dry expanded kibble diets, while high-protein or high-fiber formulations intended for dental chews, senior diets, or weight-management products may require SME inputs up to 500 kJ/kg to achieve acceptable starch gelatinization and kibble expansion. The survival loss of thermolabile vitamins, amino acids, and viable microorganisms in these systems does not depend on SME in isolation; rather, SME acts as an aggregate process variable that couples shear stress, viscous dissipation, residence time, local temperature rise, and water evaporation at the die. Field measurements on production lines show batch-to-batch survival variability of thiamine and retinol at the same apparent extruder settings, which indicates that the distribution of mechanical energy within the barrel—not solely the arithmetic mean—governs the fraction of labile molecules that survives. Because the relationship between survival loss and SME is strongly matrix-dependent, it is described through kinetic damage integrals that require knowledge of local temperature, shear rate, and residence time distributions, not through a single scalar correlation.

How Does Specific Mechanical Energy Modify Residence Time and Thermal History in a Co-Rotating Twin-Screw Extruder?

In intermeshing co-rotating twin-screw extruders, SME is generated primarily in fully filled screw elements, kneading blocks, and reverse-conveying elements, where pressure-driven backflow and inter-element leakage produce high local shear rates and viscous dissipation. The residence time distribution (RTD) in such a machine is a function of screw profile, feed rate, screw speed, and die restriction; increasing SME while holding throughput constant generally shifts the RTD toward longer mean residence times with a wider tail, although the exact behavior depends on the fill ratio. Local melt temperatures in fully filled kneading sections can rise 20 °C to 60 °C above the barrel set-point when SME is high, and these adiabatic-like excursions are often the primary cause of survival losses in heat-labile nutrients. Measurements with intrusive thermocouples on a production-scale extruder with L/D 32:1 and screw diameter 85 mm have shown that the melt temperature measured immediately before the die can differ by more than 15 °C from the barrel temperature set-point in high-SME operations. The thermal history of a nutrient-containing particle is therefore a convolution of the local temperature field and the RTD, and the cumulative heat load experienced by the fastest-moving fraction may be substantially lower than that experienced by the slowest-moving fraction. This non-uniform exposure explains why survival loss data plotted against average SME often exhibit curvature: at low SME, the fully filled zones are short and the temperature rise is modest; at high SME, the fraction of material passing through high-temperature, high-shear stagnation zones increases, and the probability of irreversible damage rises more rapidly. Steam preconditioning before the extruder further complicates the relationship because it raises entering material moisture to 20–30% wet basis and initiates partial starch gelatinization, which can reduce the torque required in the extruder and lower SME, but the added thermal load before the extruder becomes part of the cumulative survival loss if labile additives are introduced before the preconditioner. For an 85 mm diameter screw with a tip clearance of 0.3 mm rotating at 300 min⁻¹, the maximum shear rate in the clearance is approximated by γ = (π × D × N) / h, where D is 0.085 m, N is 5 s⁻¹, and h is 3 × 10⁻⁴ m, yielding a value near 4.5 × 10³ s⁻¹. Such local shear rates are sufficient to unfold proteins and disrupt encapsulated bioactives, but the fraction of total material exposed to these extreme zones is controlled by the fill ratio and screw configuration.

Thermomechanical Degradation Kinetics for Heat-Labile Vitamins and Amino Acids

For thiamine, retinol, α-tocopherol, and ascorbic acid, the accepted degradation model in low-moisture extrusion is a first-order rate expression with a temperature-dependent rate constant, extended to include shear-dependent rate enhancement. The survival fraction S is given by S = exp(−∫ k(T(t), γ̇(t)) dt), where k is the degradation rate constant, T(t) is the local temperature history, and γ̇(t) is the local shear rate history. Published kinetic studies in model low-moisture systems report activation energies for thiamine degradation on the order of 50–120 kJ/mol, with water activity and pH shifting the pre-exponential factor and the apparent activation energy. In extrusion, shear can accelerate degradation by increasing the frequency of molecular collisions, unfolding proteins, and causing direct mechanical rupture of starch–vitamin aggregates; however, the exact shear contribution is difficult to separate from simultaneous thermal effects. Retinol and α-tocopherol are additionally susceptible to oxidation, and the combination of high SME with atmospheric oxygen introduced through vent ports or low-moisture feed can promote free-radical propagation. Amino acid survival losses during companion animal extrusion are often less visible because crude protein values measured by total nitrogen do not distinguish bioavailable lysine from heat-damaged lysine. The Maillard reaction between ε-amino groups of lysine and reducing sugars is strongly temperature- and time-dependent, and elevated SME can increase the fraction of lysine that becomes nutritionally unavailable even when total lysine measured after acid hydrolysis remains unchanged. To detect this damage, feed laboratories use the difference between total lysine and reactive lysine determined by the fluorodinitrobenzene method or by O-methylisourea derivatization, with total amino acids measured according to ISO 13903:2005 after acid hydrolysis. The disparity between total and reactive lysine is a sensitive marker of thermal processing damage and tends to widen in high-SME extrusion when the same barrel temperature is maintained.

When Screw Speed and Fill Ratio Interact to Generate Non-Uniform SME Distributions

The relationship between extruder operating parameters and survival losses becomes counterintuitive when screw speed, feed rate, and moisture are adjusted simultaneously. If the screw speed is increased at constant feed rate, SME may rise because the specific energy input per kilogram increases, but the residence time may shorten, and the net effect on survival can be positive, negative, or neutral depending on the dominant damage mechanism. This interaction is governed by the fill ratio, defined as the fraction of the available screw channel volume occupied by material; fully filled zones are required for pressure generation and shear, while partially filled zones allow material to move forward with lower mechanical input. In a co-rotating twin-screw extruder, the fill ratio upstream of a restrictive element can approach 1.0, while downstream metering zones may operate at fill ratios as low as 0.2 to 0.4. High SME operation often involves a combination of low feed rate, high screw speed, long kneading block arrays, and small die openings, but these same conditions can reduce mean residence time and may therefore protect nutrients if thermal degradation dominates. Conversely, low screw speed with high feed rate can produce lower SME but longer residence time at moderate temperature, which may be equally damaging for heat-labile additives. Production-scale troubleshooting demonstrates that survival loss cannot be optimized by minimizing SME alone; the extruder must be configured to avoid simultaneous high temperature and long residence time in the presence of oxygen. For this reason, process analytical technology on modern lines includes real-time torque, melt pressure, die temperature, and in-line near-infrared moisture sensors, which allow the operator to maintain a specified SME window while monitoring the thermal load. The operational boundary for high-risk probiotic formulations is typically narrowed to a melt temperature below 70 °C and an SME below 200 kJ/kg, although published data for specific microbial inactivation constants in commercial companion animal extrusion is limited and should be verified with challenge studies on the actual formulation.

Probiotic survival in companion animal extrusion presents a distinct measurement problem because the relevant survival parameter is not a molecular concentration but a viable cell count, and the enumeration method itself can influence the reported log reduction. When dry Lactobacillus acidophilus or Bifidobacterium animalis cultures are post-extrusion coated onto the kibble surface, the survival loss during extrusion is avoided entirely; however, if the organism is incorporated before the preconditioner or extruder, the combination of steam conditioning, high shear, and die pressure can reduce viable counts by several log₁₀ cycles. The standard enumeration protocol for lactic acid bacteria in dairy matrices, ISO 20128:2006, is frequently adapted to dry pet food by homogenizing 10 g of ground kibble in 90 mL of buffered peptone water, followed by pour-plating in MRS agar at 37 °C for 72 h under anaerobic conditions. The method has an inherent lower detection limit, typically 10 CFU/g, which means that survival losses below this level cannot be distinguished from complete inactivation. Survival of probiotic strains after extrusion is also affected by the presence of protective carriers such as trehalose, skim milk powder, or maltodextrin prior to extrusion, which can shift the inactivation curve; published data for specific strain–matrix–SME combinations remains limited, and challenge testing under production-scale conditions is therefore recommended before adopting any survival claim.

Analytical Verification of Survival Losses Against AOAC and ISO Enumeration Protocols

Because survival loss is defined as the difference between the nutrient or viable cell concentration in the pre-extrusion blend and the final dried product, the analytical uncertainty of both measurements determines whether an apparent loss is statistically significant. Sampling from production-scale bulk mixers must follow ISO 6498:2012 for animal feeding stuffs, with at least three composite samples per time point to account for segregation of fine vitamin premixes. Moisture normalization is required before comparing concentrations, and moisture is determined by AOAC 930.15 using a forced-air oven. Thiamine is quantified after acid and enzymatic hydrolysis using HPLC with fluorescence detection according to AOAC 986.27 or EN 14122:2014; retinol is quantified according to AOAC 2001.13 using liquid chromatography; α-tocopherol is quantified according to AOAC 971.30 or EN 12822:2014. Total amino acids are determined according to ISO 13903:2005, while reactive lysine is measured by a separate derivatization procedure because acid hydrolysis reverses most Maillard adducts. For vitamin retention claims, the analytical coefficient of variation for duplicate samples should be below 5%, and the survival loss should exceed the expanded measurement uncertainty before process changes are considered. The following table summarizes the standard methods applicable to survival loss verification in companion animal extrusion.

Scale-up from pilot twin-screw extruders to production-scale systems requires geometric similarity of screw diameter, L/D ratio, and specific screw profile, but SME alone is insufficient as a scale-up criterion because the ratio of surface area to volume changes with diameter, altering heat transfer and the ability to remove viscous heat. In small-scale extruders with screw diameters below 50 mm, barrel cooling can mask the adiabatic temperature rise that occurs in production-scale machines with screw diameters above 100 mm, so survival loss observed at the same SME value may be lower in the pilot machine than in full production. The correct scale-up variable is not SME alone but the combination of SME, maximum melt temperature, and residence time distribution shape, all of which must be matched to preserve survival loss fidelity. Process modifications that lower SME without shortening residence time can produce the same survival loss as the original high-SME operation, and therefore any formulation change intended to improve survival should be verified across at least three production batches under normal operating variation. Post-extrusion drying in conveyor dryers at 90–120 °C for 10–30 min adds a further cumulative survival loss that is not reflected in extruder SME, so final package nutrient and probiotic counts must be sampled after drying and coating rather than immediately at the die face.