Fixed Total Particle Loading in Hybrid Nanofluids: Composition Sensitivity and Effective Property Identifiability

Fixed Total Particle Loading in Hybrid Nanofluids: Composition Sensitivity and Effective Property Identifiability

Nidhi Srivastava (1) Amit Kumar Pandey (2) Avneesh Kumar (3)

(1) Research Scholar , Dr. K.N.Modi University ,Newai Rajasthan

(2) Assitant Professor , Sharadchandra Arts Commerce & Science College Naigaon , Dist-Nanded

(3) Assitant Professor, Dr. K.N.Modi University ,Newai Rajasthan

Abstract :- Comparing hybrid nanofluids at unequal total solid loading can attribute a loading effect to composition. This analytical study fixes total volume fraction and separates the composition share from the total amount of particles. Mixture rules give exact density and volumetric heat-capacity sensitivities. A Brinkman viscosity closure predicts no composition sensitivity at fixed loading, while a clearly identified illustrative conductivity surrogate has a monotone, explicitly bounded response. The conductivity response scales with total loading in the dilute limit. A dimensionless worked example tabulates the closure outputs and demonstrates their arithmetic without representing measured AlOCu/water data. We distinguish the structural identifiability of composition from one direct conductivity observation under known constituent properties from its non-identifiability when constituent properties or transport coefficients are free. The result is a reproducible diagnostic for interpreting parametric nanofluid calculations; quantitative predictions for any material pair still require measured effective properties and independently checked closure assumptions.

Keywords: hybrid nanofluid; composition share; fixed total loading; effective conductivity; identifiability; mixture closure

1. 1. INTRODUCTION:

      Hybrid suspensions introduce at least two distinct controls: total solid volume fraction and the partition of that total between the particle species. Suresh et al. reported preparation and thermophysical properties of AlOCu/water suspensions [1], followed by an experimental heat-transfer investigation [2]. Composition and stability have also been measured at differing CuAlO mixing ratios [3]; an independent AlOCuO mixture-ratio study measured conductivity and viscosity over temperature [4]. These experiments motivate explicit control of loading, composition, temperature, and stability, but their material systems cannot be treated as interchangeable. A computation that varies loading and composition together cannot isolate a composition effect. This paper develops a fixed-loading comparison and states precisely what follows from chosen effective-property rules. Classical mixture and

      conductivity models are useful baselines [5, 6], but a model output must not be presented as a measurement.

      The analysis is deliberately restricted to algebraic closures. It does not solve a boundary-layer problem or claim heat-transfer enhancement for a physical suspension. The contribution is an exact sensitivity and identifiability audit that can be applied before a numerical flow calculation.

   2. DEFINITIONS AND PROPERTY MODEL:

      Let subscripts f, 1 and 2 denote base liquid and the two particle species. Let

      0 < < 1 be fixed and let 0 r 1 denote the share assigned to species 1:

      = + , r = 1 , = r, = (1 r) (1)

      1 2 1 2

      At fixed , ordinary volume averaging gives density and volumetric heat capacity C = (cp):

      e = (1 )f + {r1 + (1 r)2}, (2)

      Ce = (1 )Cf + {rC1 + (1 r)C2} (3)

      Mass-specific heat is cp,e = Ce/e; using a direct volume average of the mass-specific cp would be a different and generally inconsistent rule. For viscosity, adopt the Brinkman baseline [5]:

      e = f(1 )5/2 (4)

      Equation (4) is an assumption about this closure, not a universal rheological law for hybrids. It contains no r. For a transparent conductivity illustration only, first define ks(r) = rk1 + (1 r)k2 and then apply the spherical-inclusion Maxwell form:

      F(r) : = ke = ks+2kf+2(kskf) (5)

      kf ks+2kf(kskf)

      The intermediate ks is an illustrative surrogate. It does not represent a validated two-species microstructure, interfacial resistance, agglomeration, shape distribution, or percolation. Hamilton and Crosser [6] show why inclusion geometry matters even in simpler effective- conductivity descriptions. Experiments on Cu-in-ethylene-glycol nanofluids [7], proposed heat-flow mechanisms [8], interfacial-resistance theory [9], and aggregation kinetics [10]

      each identify effects absent from (5). These studies concern different systems and are cited to delimit the closure, not to validate its numerical values.

   3. EXACT SENSITIVITIES AND BOUNDS:

      Differentiation of (2)(4), holding and constituents fixed, yields

      de = ( ), dCe = (C C ), de = 0 (6)

      dr 1 2 dr

      1 2 dr

      The mass-specific heat sensitivity requires the quotient rule:

      2

      dcp,e = {(C1C2)eCe(12)} (7)

      dr e

      Write D = (1 )ks + (2 + )kf, the strictly positive denominator for positive conductivities and 0 < < 1. Direct differentiation of (5) gives

      dF = 9kf(k1k2) (8)

      dr D2

      Thus F is strictly increasing if k1 > k2, decreasing if k1 < k2, and independent of composition if k1 = k2. Its endpoint bound is min{F(0), F(1)} F(r) max{F(0), F(1)}. At small , (8) is O(): composition sensitivity necessarily vanishes as particle loading tends to zero under this closure. Likewise the density and heat-capacity differences are O().

      For k1 k2 and known kf, k1, k2, , an exact direct measurement of F in the model range gives one unique r. To make this explicit, let A = F 1 and B = F + 2. Rearrangement of (5) gives

      k = k B+2A , r = ksk2, (9)

      s f BA

      k1k2

      provided the denominator is nonzero and the recovered r lies in [0,1]. Near = 0 or k1 = k2, inversion becomes ill-conditioned or impossible. If k1 and k2 are unknown, one conductivity measurement does not separately determine them and r. A single Nusselt number additionally depends on the flow, thermal boundary conditions, and material properties, so it cannot generally identify composition without independent constraints.

      3.1 Two-parameter identifiability and measurement scale

      Suppose the total loading is also unknown, but normalized density R = e/f and conductivity F = ke/kf are both observed. Put h(r) = r1/f + (1 r)2/f, so R = 1 + [h(r) 1]. The second partial derivative of (5), in addition to (8), is

      F = 3(kskf)(ks+2kf) , D = (1 )k + (2 + )k (10)

      D2 s f

      For the map (r, ) (R, F), its local Jacobian determinant is

      .

      detJ = (12) F [h(r) 1] F (11)

      f r

      A nonzero determinant is sufficient for local recovery of (r, ) under the adopted closures and known constituent properties. It does not ensure practical precision: near a singular Jacobian, small measurement errors may produce large parameter changes. This calculation also explains why a single conductivity ratio, one scalar observation for two unknowns, cannot determine both r and .

   4. REPRODUCIBLE DIMENSIONLESS EXAMPLE

      The following ratios are selected for arithmetic illustration and are not measured properties: k1/kf = 5, k2/kf = 2, 1/f = 4, 2/f = 3, C1/Cf = 2, C2/Cf = 1.5, and = 0.03. Values below follow directly from (2)(5); e/f = (0.97)5/2 for every row.

      Table 1a. Fixed-loading compoition and conductivity inputs and outputs.

      r	1	2	ks/kf	ke/kf
      0.00	0.0000	0.0300	2.00	1.022670
      0.25	0.0075	0.0225	2.75	1.033528
      0.50	0.0150	0.0150	3.50	1.041475
      0.75	0.0225	0.0075	4.25	1.047542
      1.00	0.0300	0.0000	5.00	1.052326

      Table 1b. Fixed-loading density, heat capacity and viscosity outputs.

      r	e/f	Ce/Cf	e/f
      0.00	1.0600	1.01500	1.07912
      0.25	1.0675	1.01875	1.07912
      0.50	1.0750	1.02250	1.07912
      0.75	1.0825	1.02625	1.07912
      1.00	1.0900	1.03000	1.07912

      Figure 1. Illustrative conductivity closure at fixed total loading; the graph-paper grid is a display aid, not experimental data.

      The example has a larger ke/kf at larger r because the stipulated k1 exceeds k2. This arithmetic trend alone establishes neither a larger Nusselt number nor lower wall temperature in a flow problem. Those outcomes require a specified momentum/energy model and boundary conditions. The constant viscosity column is a useful check against a code that silently changes e when r changes under (4).

      1. Sensitivity to loading and inverse conditioning

         At r = 0.5 the derivative F/ r and the two-observable Jacobian follow directly from (8), (10), and (11). The condition number below uses the ordinary Euclidean norm of unweighted dimensionless observables; it is a numerical diagnostic, not a measurement uncertainty or a confidence interval.

         Table 2a. Conductivity ratio and composition sensitivity at r = 0.5.

         	F	F/ r
         0.01	1.013699	0.009007
         0.02	1.027523	0.018180
         0.03	1.041475	0.027522
         0.04	1.055556	0.037037
         0.05	1.069767	0.046728

         Table 2b. Loading sensitivity and inverse conditioning at r = 0.5.

         	F/	detJ	2(J)
         0.01	1.376118	-0.008757	930.0
         0.02	1.388772	-0.017675	462.8
         0.03	1.401601	-0.026758	307.1
         0.04	1.414609	-0.036008	229.2
         0.05	1.427799	-0.045430	182.5

         Figure 2. Exact local composition sensitivity at r = 0.5 under the illustrative closure.

         For example, at = 0.03, a local perturbation r = 0.02 changes F by about 0.000550, while = 0.002 changes it by about 0.002803 when varied separately. These are first- order sensitivities; the actual error budget requires observed precision, correlations, and closure uncertainty. The unweighted condition number improves as rises over this specified illustrative range, but this trend should not be generalized to arbitrary materials or measurement designs.

   5. PRACTICAL INTERPRETATION AND LIMITATIONS

      To compare two compositions fairly, record , r, the constituent data, every property closure, and the imposed thermal and mechanical boundary conditions. First compare algebraic material coefficients using (6)(8); only then attribute differences in a boundary-layer solution. An uncertainty in r propagates approximately as F (dF/dr)r locally. Uncertainty in constituent properties and model discrepancy must be considered separately; the derivative is not a confidence interval.

      The Maxwell-style surrogate neglects interactions between inclusion species. Brinkman viscosity ignores composition-dependent aggregation or rheology; observed composition stability trade-offs [3] and measured mixing-ratio effects [4] underscore this limitation. Density and heat-capacity averaging presume volume additivity. The formulas are therefore a falsifiable baseline, not a fitted description of AlOCu/water. Independently measured effective conductivity, viscosity, density, heat capacity, and particle loading are needed before quantitative material claims. Particle-migration transport would require an additional model such as Buongiornos [11]; the early nanofluid proposal [12] does not substitute for such calibration.

   6. CONCLUSIONS

At fixed total loading, composition enters density and volumetric heat capacity linearly, does not enter the chosen Brinkman viscosity, and produces a strictly monotone conductivity response in the explicitly stated surrogate when constituent conductivities differ. All algebraic composition sensitivities vanish as 0. Direct conductivity inversion is conditional on known constituent properties and the correctness of the closure. These constraints provide clear checks for subsequent hybrid nanofluid flow studies.

REFERENCES

1. S. Suresh, K. P. Venkitaraj, P. Selvakumar, and M. Chandrasekar, Synthesis of AlOCu/water hybrid nanofluids using two step method and its thermophysical properties, Colloids and Surfaces A, 388, 4148 (2011). doi:10.1016/j.colsurfa.2011.08.005.

2. S. Suresh, K. P. Venkitaraj, P. Selvakumar, and M. Chandrasekar, Effect of AlOCu/water hybrid nanofluid in heat transfer, Experimental Thermal and Fluid Science, 38, 5460 (2012). doi:10.1016/j.expthermflusci.2011.11.007.

3. F. R. Siddiqui, C. Y. Tso, K. C. Chan, S. C. Fu, and C. Y. H. Chao, On trade-off for dispersion stability and thermal transport of Cu AlO hybrid nanofluid for various mixing ratios, International Journal of Heat and Mass Transfer, 132, 12001216 (2019). doi:10.1016/j.ijheatmasstransfer.2018.12.094.

4. V. V. Wanatasanappan, M. Z. Abdullah, and P. Gunnasegaran, Thermophysical properties of AlOCuO hybrid nanofluid at different nanoparticle mixture ratio: An experimental approach, Journal of Molecular Liquids, 313, 113458 (2020). doi:10.1016/j.molliq.2020.113458.

5. H. C. Brinkman, The viscosity of concentrated suspensions and solutions, Journal of Chemical Physics, 20, 571 (1952). doi:10.1063/1.1700493.

6. R. L. Hamilton and O. K. Crosser, Thermal conductivity of heterogeneous two-component systems, Industrial & Engineering Chemistry Fundamentals, 1, 187191 (1962). doi:10.1021/i160003a005.

7. J. A. Eastman, S. U. S. Choi, S. Li, W. Yu, and L. J. Thompson, Anomalously increased effective thermal conductivities of ethylene glycol-based nanofluids containing copper nanoparticles, Applied Physics Letters, 78, 718720 (2001). doi:10.1063/1.1341218.

8. P. Keblinski, S. R. Phillpot, S. U. S. Choi, and J. A. Eastman, Mechanisms of heat flow in suspensions of nano-sized prticles (nanofluids), International Journal of Heat and Mass Transfer, 45, 855863 (2002). doi:10.1016/S0017-9310(01)00175-2.

9. C.-W. Nan, R. Birringer, D. R. Clarke, and H. Gleiter, Effective thermal conductivity of particulate composites with interfacial thermal resistance, Journal of Applied Physics, 81, 66926699 (1997). doi:10.1063/1.365209.

10. R. Prasher, P. E. Phelan, and P. Bhattacharya, Effect of aggregation kinetics on the thermal conductivity of nanoscale colloidal solutions (nanofluid), Nano Letters, 6, 15291534 (2006). doi:10.1021/nl060992s.

11. J. Buongiorno, Convective transport in nanofluids, Journal of Heat Transfer, 128, 240250 (2006). doi:10.1115/1.2150834.

12. S. U. S. Choi and J. A. Eastman, Enhancing thermal conductivity of fluids with nanoparticles, Argonne National Laboratory report ANL/MSD/CP-84938 (1995).