Artificial intelligence driven multiphysics analysis of Casson bioconvective ternary nanofluid flow over an inclined needle Authors: Amir Abbas, Soubhagya Kumar Sahoo, Muhammad Farman, Soumyarani Mishra, Aboulbaba Eladeb, Nermeen Abdullah, Walid Hassen, Lioua Kolsi Published: 2026 ABSTRACT AbstractThis study investigates the bio-convective flow of ternary Casson nanofluid across inclined moving thin needle embedded in a porous medium, considering Arrhenius activation energy and non-uniform heat generation/absorption. Such models are important in biomedical engineering, thermal energy systems, and advanced cooling technologies, where enhanced heat and mass transfer characteristics are required. The governing equations are solved numerically using the bvp4c technique, while Artificial Neural Network (ANN) analysis is employed to evaluate the accuracy and stability of the computational results. The findings reflect that increasing Casson parameter, activation energy parameter, needle thickness, bioconvection Lewis number, Peclet number, and chemical reaction parameter led to reduction in velocity distribution and reverse trend for increasing inclination angle of surface. Results show that rising chemical reaction parameter, activation energy parameter led to rise temperature distribution and opposite behavior for needle thickness. It has been reflected that increasing activation energy parameter caused to increase concentration. Graphical outcomes show that microorganism density rises with the rise of activation energy parameter, bioconvection Lewis number, Peclet number, and inclination angle. Furthermore, the ANN model demonstrates high predictive accuracy with small and uniformly distributed errors, confirming the reliability of the numerical analysis. For the better visibility, three-dimensional graphs of the velocity, temperature and concentration and microorganisms’ density profiles are shown. For compact visualization of the behavior of velocity and temperature, the graphical representation of contours plots are also presented Current results are validated with published results in the literature.Keywords: Ternary nanofluid, Arrhenius, Casson fluid, Inclined needle, Non-uniform heat rise/fall Introduction The collective, frequently random movement of motile microorganisms that produces fluid motion and instability due to their asymmetric mass distribution and their responses to factors such as gravity, chemical gradients, or magnetic fields is known as bioconvection. Microorganisms that move against gravity, such as gyrotactic bacteria or algae, can cause hydrodynamic instability and convection patterns in bio-convective systems. Because of these characteristics, bioconvection is a key component of bio microsystems and has potential applications in biotechnology, fuel cells, photobioreactors, enzyme biosensors, and ecological energy systems. Bioconvection has been extensively investigated through theoretical, numerical, and experimental approaches. The phenomenon was illustrated using a floating algae suspension [1], followed by the pioneering theoretical model of Plesset and Winet [2], which incorporated a wide range of motile microorganisms. Subsequently, Kuznetsov [3] developed a computational model demonstrating that cell deposition plays a crucial role in enhancing bioconvection growth. With increasing interest in nanofluid systems, several studies have examined the interaction between motile microorganisms and nanoparticles under various physical conditions. Waqas et al. [4] analyzed bioconvection through electrically conducting nanofluid flow across porous stretching disk, while Khan and Shehzad [5] explored bioconvection in Carreau nanofluid along stretching surface. Effects of radiation and porous media on oxytactic bacterial bioconvection were reported by Balla et al. [6]. Basir with his research fellows [7] considered Peclet number’s influence on nanofluid flow across stretching cylinder. Beyond Newtonian nanofluids, bioconvection studies have been extended to non-Newtonian and complex fluid models. Kuznetsov [8] reported oscillatory bioconvection in shallow nanoparticle layers containing oxytactic microorganisms. Uddin et al. [9] examined bioconvection in non-Newtonian fluids within saturated porous space and later investigated effects of slip and blowing over horizontal surfaces [10]. Rashad and Nabwey [11] analyzed mixed convection nanofluid flow with motile microorganisms under convective boundary conditions. Additional studies addressed variable viscosity effects in rotating bioconvective systems [12], bioconvection in horizontal channels using modified least-squares techniques [13], slip mechanisms induced by gyrotactic microorganisms [14], and the rheological behavior of Oldroyd-B nanofluids over oscillatory stretching surfaces [15]. More recently, Waqas et al. [16] quantitatively examined bioconvection in rate-type (Maxwell) fluid flows using the shooting method. Recent investigations have incorporated chemical reactions and thermodynamic irreversibility into bioconvection models. Khan et al. [17] analytically considered effect of chemical reactions on the peristaltic bioconvection flow of a third-grade nanofluid. In [18], transport characteristics of Casson hybrid blood flow containing oxytactic bacteria in a stenosed artery were analyzed using Arrhenius activation energy. Furthermore, Mondal et al. [19] investigated hydromagnetic bioconvective nanofluid flow incorporating solar radiation, viscous and Joule dissipation, activation energy, Brownian motion, Hall current, and thermophoresis, and reported that buoyancy enhances velocity while entropy generation increases due to radiative and dissipative effects. Casson fluid is shear-thinning non-Newtonian fluid, in which shear stress causes its viscosity to drop. This model has established a lot of interest recently in field of Nanofluid research, especially when it comes to modelling biological fluid dynamics like blood flow. Because of its significance in these fields, mathematicians and researchers are focusing on it to expand knowledge across the theoretical and practical sciences. Nanofluids are formed by combining nanoparticles into base fluids, are the result of the evolution of nanotechnology. Because of their improved thermal properties, these nanofluids have outperformed conventional fluids. Due to their exceptional thermal performance and several potential uses in engineering systems, hybridized ternary nanofluids, made of three kinds of nanoparticles, have caught a lot of attention lately. By utilizing the combined effects of nanoparticles, ternary nanofluids provide important developments in overall thermal transfer and performance. The potential for this technology to transform heat regulation for industrial applications makes it a promising idea in thermal engineering. Magnetohydrodynamics slip flow of ternary hybrid Casson fluids across nonlinear stretching surfaces embedded in porous media is key topic of research. Ali et al. [20] focused on the significance of applied magnetic fields and convective barriers in enhancing heat transfer efficiency. Mahanta and Sharma [21] explained how thermal conductivity and energy profiles have been improved by the ternary hybrid nanofluids. Jat et al. [22] directed computational analyses to study flow features of Casson fluids in such setups, underscoring their relevance in optimizing industrial processes. These studies incorporate advanced concepts such as Brownian motion, thermophoresis, and heat transfer law in generalized form to provide extra realistic representation of real-world scenarios offered by Lone et al. [23]. Li et al. [24], in their article, studied stimulus of induced magnetic fields on flow process of thin Casson fluid films, whereas Bixapathi and Babu [25] investigated the stability of buoyant flows under magnetic fields. Sharif et al. [26] investigations demonstrated that as Casson and magnetic parameters increase, Bejan number and entropy generation show differing behavior, providing important information for thermal system optimization. Additionally, Casson fluids are being used in scientific domains like blood flow modeling extensively. Samal et al. [27] inspected sensitivity of thermal transfer and the development of entropy in the electroosmotic bioconvection of blood-based Casson hybrid nanofluids in squeezing channels. The flow of Casson fluids through vertical microchannels under impacts of radiation and magnetic force have been studied by Sreehari et al. [28], with an attention on entropy generation and heat transfer optimization. Lee [29], who first examined boundary-layer development across thin needle and established theoretical framework for subsequent studies, study of boundary-layer flow across thin-needle geometries has evolved gradually over several decades. The understanding of thermal transport mechanisms in slender bodies was greatly enhanced by the groundbreaking work of Narain and Uberoi [30–32], who systematically investigated forced, free, and mixed convection from thin needles. Patil et al. [33] studied Casson-Williamson ternary nanofluid past the yawed. In [34], the Sakiadis convective flow for hybrid nanofluid flow at a moving thin needle is discussed. Impact of thin needle and velocity ratio on the properties is determined. Khan et al. [35] accomplished numerical simulation of bio-convective micropolar nanofluid flow past moving needle under effect of Arrhenius activation energy, and binary chemical reaction and dissipation control. The study of heat and mass transmission in two phase model of the nanofluid taking into effect of thermophoresis and Brownian motion across permeable thin needle along with sensitivity analysis has been reported in [36]. Nehaal et al. [37] discussed ternary nanofluid flow past diverse geometries with chemical reaction, magnetic force, heat source and sink along with activation energy, joule heating and dissipation impacts. In Ref [38]. examined nanofluid flow in Krieger-Dougherty and Maxwell-Bruggeman models with thermal slip conditions of velocity slip conditions along with Hull current in porous space. Iqbal et al. [39] described analysis of entropy generation analysis in hybrid nanofluid incorporating dissipation effects, velocity slip, Ohmic heating, solar radiation, and dissipation impacts in porous space along with slip conditions. Iqbal with his co-research scientist [40] analyzed an influence of ternary nanofluid flow and heat transmission process past curved stretching surfaces. Farooq et al. [41] scrutinized unsteady flow of hybrid nanofluid past spinning disk in the porous space with magnetic force, Coriolis force, and dissipation, solar radiation, velocity slip. Bioconvective flow micropolar nanofluid in control of thermal diffusion and diffusion thermo phenomenon computing non-similar solutions has been highlighted in [42]. Varatharaj et al. [43] explored ternary nanofluid flow along permeable sheet with slip conditions. Ishaq et al. [44] scrutinized on the boundary layer flow over the continuously moving thin needle. Biological neural networks present in human brain serve as basis for simulation models known as artificial neural networks (ANNs). They contain of layers of organized nodes or neurons, including an input layer, one or more hidden layers, and an output layer. It may learn and model complex correlations between inputs and outputs through data training. When it comes to handling large amounts of data and non-linear relations, they are significantly better than other computing methods. Without requirement for explicit characteristics engineering, which is sometimes a laborious and challenging process in conventional methods, they are able to automatically extract features and patterns from raw data. Ziaur et al. [45] carried out the analysis of artificial neural network prediction for free convection in a V-Shaped geometry whose bottom wall is filled with nanofluid. Nihaal et al. [46] discussed a flow of bioconvection in nanofluid incorporating chemical reaction and heat source and sink influences past thin needle under non-uniform heat rise and fall effects using stochastic approach They carried out artificial neural network analysis as well. A numerical analysis of artificial neural network based Casson hybrid nanofluid flow and energy transmission on porous sheet has been carried out in [47]. Akbar et al. [48] did the analysis of carried out the artificial neural network-based study of thermal mass transport through nanofluid peristaltic flow in porous media. Akbar et al. [49] carried numerical simulation using artificial neural network of the thermal transportation in parabolic trough solar collectors using hybrid nanofluids flow with incorporation of solar radiation, generation of heat, Darcy-porous media, and heat dissipation impacts. Farooq et al. [50] carried out artificial neural network analysis for Eyring-Powell nanofluids with magnetic force, Darcy porous media, dissipation impacts, solar radiation and Joule heating control. Alharbi and Usman [51] inspected artificial neural network analysis of three-dimensional flow of ternary nanofluid across stretching sheet with incorporation of solar radiation, generalized heat transfer law, transpiration, inclined magnetic force and convective boundary conditions. Yaseen et al. [52] inspected ternary nanofluid flow in rotating horizontal annulus between two coaxial tubes using artificial neural network under control of shapes of nanoparticles, dissipation, solar radiation. Turning the pages of the published literature, it has been concluded that many studies separately discussed on the microorganism flow of nanofluids such as mono-nanofluid, hybrid nanofluid, ternary nanofluid with magnetohydrodynamic, dissipation effects, coupled in non-Newtonian nanofluid along the different geometries including thin needle. However, the combined impact of activation energy, Lorentz force, Brownian motion, thermophoresis and viscous dissipation in Casson ternary fluid along the moving thin needle embedded in a porous media. Non-uniform heat rise and fall effects are also taken into consideration. The model further incorporates gyrotactic microorganisms to analyze bioconvection behavior and nanoparticle stability, which has not been extensively explored for ternary nanofluids in thin needle geometries. To enhance heat transfer and flow management in industrial and biomedical contexts, their research examines the interactions among magnetic fields, porous media, and thermal radiation. Our theoretical knowledge of non-Newtonian fluid mechanics is enhanced by the results of these studies, which also open the door for creative engineering solutions in various fields, such as energy systems and medical devices. Mathematical formulation Consider, viscous, two-dimensional, incompressible fluid flow of Casson ternary nanofluid. Lorentz force is employed in vertical direction of flow. Non-uniform heat rise and fall are taken into account. The thin needle is embedded in the porous medium. Casson fluid model is considered in current study. A single-phase model of nonfluid is used, considering /water based ternary nanofluid. The microorganism density is considered. The Arrhenius activation effects are considered. The whole flow structure is given in Fig. 1. Succeeding [33–35, 43–44], the proposed equations are presented below: Flow structure The corresponding boundary conditions are: Symbols appeared in above equations are velocity in and direction or axial and radial velocity components , porosity of porous medium , non-uniform heat source parameter , Concentration in the fluid , non-uniform heat volumetric heat generation/absorption term (), density , non-uniform heat sink parameter , velocity field , specific heat coefficient , Maximum cell swimming speed , temperature of the fluid at wall , viscosity coefficient , reaction order (), temperature of fluid , gravity , chemical reaction rate constant , Brownian diffusion , activation energy , molecular diffusivity , magnetic field strength , environment concentration , ratio of nanoparticle heat capacity to base fluid heat capacity , environment temperature , concentration expansion coefficient , electrical conductivity , thermophoretic diffusion coefficient , microorganism density , thermal conductivity , and Inclination angle of the surface . Following [34], the non-uniform heat source/sink in energy expression is denoted as and mathematically can be stated as . The physical properties in the form of gradient like , , ,and , the skin friction coefficient, Nusselt number, Sherwood number, and Microorganism’s density gradient are defined in the following equation as given in [34–35], The thermophysical properties considered in the current problem are defined in Table 1. Correlations of thermophysical properties for ternary-hybrid nanofluids are given in Table 2. Table 1Nanoparticle’s properties [33]Properties Cp (J Kg-1 K-1) k (W m-1 K-1) (S/m) β × 10-5 K-1 Base Fluid (Water)997.1 ρ (Kg/m)41790.6135.5 x 10-62.1 x10-4Alumina3970765403.5 x 1060.85 x 10-5Titania4250686.28.952.83 x 1060.9 x 10-5Silicon dioxide (22007451.43.5 x 1060.9 x 10-5 Nanoparticle’s properties [33] Table 2Properties correlations for ternary-hybrid nanofluids [43]PropertiesTernary nanofluid (thnf) & hybrid nanofluid (hnf)Density Heatcapacity Dynamic viscosity Thermal expansion coefficient Thermal conductivity Electrical conductivity Properties correlations for ternary-hybrid nanofluids [43] Heat capacity Here, , and represent volume fractions of Alumina Titania, and Silicon dioxide ( respectively. Here, 1, s1, s2, s3 represents base fluid, nanoparticles (), nanoparticles (), and nanoparticles (), respectively. Solution methodology The differential Eqs. (1–5) along the boundary conditions (6) are transmuted into ordinary differential equations by means of very suitable similarity variables given in Eq. (8). After the conversion, obtained set of equations is solved using bvp4c solver. Similarity analysis For the conversion from a set of PDEs to a set of ODEs, the variables for similar solutions defined in Eq. (8) by following [34] are used. The concerned similarity variables are: where, are non-dimensional stream function, composite velocity, and similarity variable. The velocity component in terms of dimensionless stream function has been defined as . Using variables given in Eq. (8) into Eqs. (1–6) we have following transformed Eq. where, Here, dimensionless parameters are magnetic field parameter Casson fluid flow parameter , Darcy number , buoyancy ratio parameter , mixed convection parameter , dimensionless microorganism density , Rayleigh number , Prandtl number , thermophoresis parameter , bioconvection Lewis number , velocity ratio , Chemical reaction parameter , needle thickness , Peclet number , temperature ratio parameter Activation energy parameter , Reynolds number Schmidt number , Eckert number , and Brownian motion parameter . The transformed form of Eq. (7) is given below: Solution technique For the simulation of transformed Eqs. (9–14), the bvp4c, a MATLAB solver is used. To apply the MATLAB solver, the Eqs. (9–13) are converted to system of first-order ordinary differential equations and lastly solved by means of bvp4c numerical algorithm. The reduced set of first order ordinary differential equations are given below for the current problem. BCs, The transformed set of first order ODE given in Eqs. (15–20) are inserted into built-in numerical solver bvp4c in MATLAB software for computations of physical quantities like velocity, temperature, concentration and microorganisms’ profiles along with their gradients in form of graphs and tables. The iteration continues till criteria of convergence is met i.e. . The process satisfies the boundary conditions in asymptotic manner. Using a 20 by 20 mesh size, convergence conditions for at each step are displayed below: An iteration process begins at and ends at , which is the infinity value. Excellent asymptotic solutions are confirmed to satisfy the problem boundary condition according to the convergence criterion indicated in Eq. (21). This pattern shows that BCs are satisfied in terms of mesh size and computation. CPU run time is 15 s for three different parameter values, demonstrating the effectiveness of the computation used in the system. In the next section, the results and discussion are elaborated upon. Results and discussion numerical boundary value problem analysis using bvp4c In the current section, the influence of the physical number on the quantities of interest, that is, velocity, temperature, concentration, and microorganism profiles, will be discussed in detail. Based on physical reasoning, the graphs and tables will be elaborated upon and concluded. Effect of pertinent parameters on the velocity field Graphical fallouts for the velocity profile versus Casson fluid parameter are shown in Fig. 2. The results show that monotonically decreases when increases for simple fluid, mono, hybrid, and ternary nanofluids. Since Casson fluid parameter represents the yield stress of a non-Newtonian fluid, when the values of increase, the resistance to the deformation, as well as the high values of apparent viscosity, are secured. Hence, a greater force is required to initiate and sustain motion; therefore, this weakens the flow field, that causes drop in velocity field throughout boundary layer. From graph, overall, it can be concluded that the decreasing trend with rising magnitudes is due to the dominant role of yield stress and viscous resistance. The impact of on is shown in Fig. 3. The parameter symbolises the activation energy parameter that shows up in the concentration equation’s Arrhenius reaction term. The chemical reaction rate decreases as increases because of the exponential suppression. As a result, there are more species present in the boundary layer. The increased concentration exacerbates the opposite density gradient, which suppresses the flow field since concentration is connected to the momentum equation via buoyancy forces. For all fluid configurations, the velocity profile thereby drops as increases. For all fluid types, it is gotten that increasing reasons velocity to drop monotonically. Higher values of physically intensify the non-Newtonian material effects, thereby restricting momentum transmission and increasing internal resistance to deformation. Hence, momentum boundary layer thickness decreases, and the velocity field weakens. Furthermore, because of its somewhat lower effective viscosity, the simple fluid has the maximum velocity, while the ternary nanofluid has the lowest, followed by the hybrid and mono-nanofluids. All things considered, an increase in increases viscous resistance, thereby reducing the strength of the boundary layer. Figure 4 presents the influence of needle thickness on velocity profile . The velocity cuts with swelling values of ‘c’ for all fluid configurations. Physically, increasing ‘c’ enhances the concentration level within the boundary layer. Since concentration is coupled with the momentum equation through the solute buoyancy term, higher needle thickness strengthens the opposing density gradient, which suppresses fluid movement. Subsequently, momentum boundary layer thickness reduces and the velocity profile declines. Moreover, ternary nanofluid displays the lowest velocity due to its higher effective density and viscosity, while simple fluid maintains highest velocity. Overall, a stronger needle thickness weakens the flow field. Display of the versus Display of the versus Display of the versus Figure 5 demonstrates how velocity profile is controlled by bioconvection Lewis number . For all fluid types, it is found that velocity falls as rises. Physically, lower microbe diffusivity is correlated with bigger values of , which promotes microorganism concentration close to the surface. The greater accumulation increases the opposing buoyancy force, which suppresses fluid motion since the density of microorganisms is included in the buoyancy part of the momentum equation. Hence, velocity profile shrinks and the momentum boundary layer thins. Because of its higher effective viscosity and density, the ternary nanofluid has the lowest velocity, whereas the simple fluid has the highest velocity. In general, the flow field is weakened by increased bioconvective processes. Control of Peclet number on velocity profile has been depicted in Fig. 6. For all fluid configurations, it is found that the velocity falls as increases. Stronger advective transport in relation to microorganism diffusion is physically indicated by a greater . This intensifies the bioconvective buoyancy factor in the momentum equation by increasing the collection of microorganisms within the boundary layer. A thinner momentum boundary barrier and decreased velocity are the results of fluid motion being suppressed by the increased opposing buoyancy force. Among the fluids, ternary nanofluid exhibits lowest velocity owing to its higher effective viscosity and density, while simple fluid maintains highest velocity. In general, dominating convective bioconvection forces cause the flow field to diminish as increases. Figure 7 reveals effect of inclination angle on the velocity profile . Velocity surges with increasing values of . The velocity profile increases with larger values of inclination angle α because the component of buoyancy force acting along flow direction becomes stronger. As α increases, the gravitational force contributes more effectively to the fluid motion parallel to the surface, which enhances the momentum of fluid particles. Consequently, the resistance to flow decreases and the fluid accelerates, leading to higher velocity distributions in boundary layer. Graphical results for velocity profile versus the chemical reaction parameter are shown in Fig. 8. The results show that monotonically decreases when increases for simple fluid, mono nanofluid, hybrid nanofluid, and ternary nanofluid. This behavior proposes that the parameter introduces a resistive effect in the flow, thereby conquering the fluid motion across all fluid configurations. When increases then strong chemical reaction accurs that minimizes the concentration of diffusion species in boundary layer. This causes to reduce solutal buoyancy force which is one of the most important driving forces of the flow. Consequently, movement of fluid is slowed down, leading to decrease velocity distribution in entire boundary layer. Display of the versus Display of the versus Display of the versus Display of the versus Influence of pertinent parameters on the temperature field Effect of on is publicized in Fig. 9. For all fluid configurations, it is found that the temperature rises as grows, resulting in a heavier thermal boundary layer. Larger physically alters fluid’s thermal transport mechanism and increases internal resistance. As a result, the temperature distribution in boundary layer region is raised since rate of heat diffusion away from the surface is decreased. Because of its increased effective thermal conductivity and heat capacity, the ternary nanofluid has relatively higher temperature levels among the fluids, whereas the simple fluid has the lowest temperature profile. In general, the flow domain retains more thermal energy as increases. Figure 10 displays how temperature profile is exaggerated by activation energy parameter . It is found that for all fluid arrangements, the temperature rises somewhat as grows. Because of the exponential Arrhenius term, a higher activation energy physically reduces the rate of chemical reactions. Reducing species consumption and changing energy exchange within the boundary layer, the weakened reaction improves thermal energy retention. This results in a slight increase in the thickness of thermal boundary layer. The basic fluid has the lowest temperature distribution between the fluids, while ternary nanofluid has a somewhat advanced temperature because of its improved effective thermal characteristics. In general, the accumulation of thermal energy within the flow domain is encouraged by raising . Control of needle thickness ‘c’ on temperature distribution is presented in Fig. 11. For all fluid types, it was found that raising ‘c’ marginally lowers the temperature profile. Physically, the higher the needle thickness, the faster the reaction, increasing species consumption and altering the boundary layer’s energy balance. So, thermal energy decreases, and thickness of thermal boundary layer decreases. It improved effective thermal conductivity, the ternary nanofluid has a somewhat higher temperature than the other fluids, while the simple fluid () has the lowest temperature distribution. The general pattern demonstrates that temperature field in flow domain is suppressed by the stronger needle thickness. Display of the versus Display of the versus Display of the versus Influence of pertinent parameters on the concentration field The consequence of activation energy parameter on concentration is shown in Fig. 12. For all fluid models, it is found that improves the concentration profile. Higher activation energy physically reduces species consumption by weakening the chemical reaction rate because of the Arrhenius exponential term. Consequently, the concentration rises, and the concentration boundary layer thickens because more solute particles stay in the boundary layer. Additionally, simple fluid has bottom profile, while ternary nanofluid has the largest concentration dispersion, tracked by hybrid and mono nanofluids. Such behavior is linked to altered thermophysical and diffusivity characteristics in suspensions of many nanoparticles. Larger generally reduces the strength of chemical reactions, which raises the concentration field across the flow domain. Display of the versus Influence of pertinent parameters on the microorganism’s concentration field Figure 13 portrays the influence of activation energy parameter on the dimensionless microorganism density profile . It is observed that increasing leads to an enhancement in the concentration distribution within the boundary layer, convoyed by thickening of concentration boundary layer of microorganisms. Materially, higher activation energy increases the energy barrier of the chemical reaction term , which weakens the reaction rate and reduces species consumption. As a result, fewer reactants are depleted, leading to higher concentration levels throughout the boundary layer. Consequently, mass diffusion dominates over chemical reaction effects, allowing microorganism density profile to persist further into the flow domain. Overall, larger values of suppress chemical reaction strength and enhance microorganism density profile accumulation, strengthening the concentration field within the system. Display of the versus Stimulus of bioconvection Lewis number on dimensionless microorganism profile . for various fluid configurations is depicted in Fig. 14. Increasing values of leads an enhancement in . This happens epitomizes ratio of thermal diffusivity to microorganism diffusivity. Higher values lead to reduction in microorganism’s diffusivity, it is meant by it that slow diffusion occurs in boundary layer. This reduction compels to collect near a sheet instead of spreading to entire the boundary layer. In consequent to this microorganism intensification leads to increased microorganisms’ density profile. Display of the versus Impact of Peclet number on the dimensionless microorganism density profile for various fluid configurations is shown in Fig. 15. Increasing is found to greatly increase the concentration of microorganisms close to the surface, which then gradually decreases throughout the boundary layer. Physically, convective transport predominates over diffusive transport, which increases the advection of microorganisms toward the surface and encourages their aggregation in the near-wall zone. This is shown by a greater Peclet number. Subsequently, bioconvective boundary layer thickens and peak value of rises. Due to increased viscosity resistance and decreased microbe mobility brought on by multiple nanoparticle inclusions, the ternary nanofluid has the lowest dispersion amongst fluid types, whereas simple fluid has highest microorganism density. Overall, increasingstrengthens convective transport, intensifies microorganism accumulation near the surface, and enhances bioconvection effects within the flow domain. Effect of inclination angle on the dimensionless microorganism density profile for various fluid configurations is shown in Fig. 16. As increases, the peak value of increases and the spread of microorganisms within the border layer is generally suppressed. Physically, the inclination angle is represented by the quantity in the buoyancy term; as increases, effective component of buoyancy force acting along flow direction decreases. This reduces the concentration of microorganisms close to the surface and increases bioconvective movement. Due to increased viscosity resistance and decreased microbe movement brought on by multiple nanoparticle inclusions, the ternary nanofluid shows the lowest dispersion among the fluid types, whereas the simple fluid retains the maximum microbial density. Overall, increasing the inclination angle, increases microorganism accumulation, and enhances bioconvective boundary layer development. Display of the versus Display of the versus Influence of pertinent parameters on the skin friction coefficient Impact of pertinent parameters like , , , and on the skin friction coefficient will be shown in Figs. 17, 18, 19 and 20, respectively. Figure 17 shows that when the Casson fluid flow parameter increases, then gradually decreases for all of the cases of the fluid that is a simple, nanofluid, hybrid nanofluid, and ternary nanofluid. These occurring phenomena tells that a reduction in skin friction coefficient due to increasing Casson fluid parameter leads to attenuation in effective viscous resistance in fluid. The perceptible point is that extreme magnitude for a simple fluid is secured. Figure 18 shows the influence of on , it can be observed that as rises skin friction decreases. From physics view point, strong chemical reaction declines species concentration that contribute in mass transportation inside boundary layer. This makes weaker the buoyancy term due to solute that minimizes momentum transportation near the wall. Therefore, velocity gradient near the wall gets weaker compelling to reduction in wall shear stress. As there is direct relation between and wall shear stress so skin friction shrinks with augmenting . The influence of needle thickness on has been shown in in Fig. 19. Graphs shows that when rises, diminishes. The maximum curve for the is obtained for a simple fluid. Figure 19 shows the influence of on , it can be detected that as rises the skin friction coefficient decreases for all cases. The maximum curve for the is obtained for simple fluid. When needle thickness increases it becomes thicker then contact surface area increases and then it reduces gradient of the velocity near the surface. Due to thickness of the needle feels lower shear deformation that causes to get attenuated the shear stress at wall. There is direct dependance of on shear stress and so it slowly decreases when increases. Figure 20 shows the influence of on , it can be detected that as rises skin friction coefficient decreases for all of the cases. The maximum curve for the is obtained for simple fluid. As is increases then diffusion of microorganism is smaller than momentum diffusion and thermal diffusion due to which small microorganisms’ distribution is small in the boundary layer. This confinement reduces gradient of density, hence, diminishes buoyancy driven increment in fluid movement near surface. This is happening for call cases of nanofluids. Display of the versus Display of the versus Display of the versus Display of the versus Influence of pertinent parameters on the Nusselt number Effect of the appropriate parameters like , on the Nusselt number are shown in Figs. 21 and 22. Figure 21 depicts declines when is augmented for all cases, and the highest value is achieved for ternary nanofluid. The effect of needle thickness on is depicted in Fig. 22. Figure 22 reflects that when is raised then get weaker for all of the types of fluids in the current case. Display of the versus Display of the versus Influence of relevant parameters on 3D graphs of velocity profile, temperature profile, concentration profile, and microorganisms’ density. Figure 23a presents a three-dimensional depiction illustrating the interaction among the velocity field, magnetic field parameter, and the similarity variable η, thereby clarifying the manner in which fluid velocity develops within the system. Figure 23b demonstrates the dynamic variation of the temperature distribution, providing an extensive representation of the thermal behavior over all fluids. Figure 23c portrays progression of the concentration field, highlighting the variation in species distribution throughout the flow regime. These graphical illustrations significantly enhance the interpretation of mass transport mechanisms and dispersion characteristics. Furthermore, Fig. 23d illustrates the evolution of microorganism density, offering additional insight into the bioconvective transport processes within the system. When increases then Lorentz force increases then resistance in the fluid rises hence velocity of the fluid declines and temperature of fluid rises exact at surface. Display of the 3D plots of a b θ, versus Influence of relevant parameters on contours plots of velocity profile, temperature profile, concentration profile, and microorganisms’ density. The contour plots of are shown in Fig. 24a–d for M = 0.1. A comparatively low Lorentz force is applied; therefore, a mild impact on the flow. The contours illustrate the coupled behavior of velocity, temperature, concentration, and microorganism density within the boundary layer. Figure 24a shows that velocity is extreme near wall and cuts smoothly as surges, impending the free-stream condition. The contour lines are moderately broadly spaced away from the wall, representing steady momentum diffusion. Since (M = 0.1) is minor, the resistive Lorentz force is feeble, so the fluid experiences negligible magnetic braking. The boundary layer thickness is consequently moderately greater than it would be at higher (M), because the flow is not powerfully repressed. A small magnetic parameter implies weak electromagnetic resistance. Hence, the fluid retains more kinetic energy, follow-on in a thicker momentum boundary layer and smoother velocity gradients. Figure 24b characterizes the temperature contours. Temperature is maximum close to surface and decreases quickly with increasing depth Closely packed contours near the wall designate a robust thermal gradient. The thermal boundary layer is thinner compared to the momentum layer. Heat transfer is mainly ruled by conduction near the wall and diffusion outward. Since magnetic heating effects are weak at (M = 0.1), temperature declines gradually, forming a different but reasonable thermal boundary layer. Figure 24c represents the contours of the concentration profile. It is maximum close to wall and reduces far from it. The contour spacing proposes steady mass diffusion. Mass diffusion is normally slower than heat diffusion (depending on the Schmidt number). With a feeble magnetic effect, species transport remains dominated by diffusion and convection, resulting in a smooth concentration gradient. Figure 24d reflects an effect of magnetic field on the contours of the microorganism density. The contours near the surface are closely clustered, reflecting that there is a large microorganism density gradient. The distribution reflects bioconvection effects influenced by fluid motion. Microorganism transport depends on both fluid velocity and concentration gradients. Since the magnetic field is weak, the flow supports microorganism movement without strong suppression, leading to a well-defined bioconvective boundary layer. Figures 25a-d represent contours of the velocity, temperature, concentration and microorganisms’ density profiles for when the rest of the parameter conditions are kept fixed. The enhanced values of the creates larger Lorentz force; therefore, fluid velocity is slowed down as shown in the contour plots in Fig. 25a. Similarly, the Lorentz force decreases velocity, because it improves interior energy generation due to Joule heating. Reduced convection, combined with additional heat production, causes the temperature to diffuse further into the fluid. So, the thermal boundary layer thickens for higher . Since velocity is blocked, convective mass transport deteriorates. With reduced fluid motion, species diffusion controls, allowing concentration to continue beyond the wall. Therefore, the concentration boundary layer develops more thickly at an advanced magnetic parameter. The phenomenon is shown in Fig. 25c. Motile microorganism density contours ξ(η), it spreads into the flow domain. The spacing in the contours indicates that a thick bioconvection boundary layer is developed. Close to the wall gradients decreases comparison the small values of . The magnetic force decreases the fluid movement and microorganism transport due to convection diminishing. The motion of microorganisms depends on fluid velocity and concentration gradients. Consequently, microorganisms accumulate and diffuse more gradually, increasing the thickness of the microorganism distribution layer. a-d Display of Contours plots a for a-d Display of Contours plots a for Artificial neural networking (ANN) Artificial neural networks (ANNs) are computational models that are used to find intricate, nonlinear relationships between inputs and outputs. Several fields in which it is used for modelling in finance, environmental science, forecasting, and optimization. There is a classification of the ANN, such as single-layer, multi-layer, feed-forward, or recurrent. This flexibility shortens their use in a range of complex employment. There are the input (magnetic field parameter and similarity variable parameter) and output (velocity profile and concentration profile). An ANN model was created using MATLAB’s Neural Network Fitting Tool (nftool). It regularly standardizes all input and output data in using built-in map min-max function. Figure 26 demonstrates a simple feed-forward neural network consisting of three main components: an input layer, a hidden layer, and an output layer. Figure 27a–f assesses the train ANN model of velocity profile; here, and are input parameters and velocity profile is out parameter. Figure 27a shows the excellent agreement between the bvp4c and ANN approximation. Figure 27b shows the best validation performance of mean square error at 4.7443 and epoch 43. Figure 27c shows the validation check at epoch 43. Figure 27d shows the Error Histogram of Neural Network Performance for Training, Validation, and Test Sets. This is commonly used for performance of the Artificial Neural Network. This figure in a histogram represents the difference between the predicted output and the target (true) value. The values near the zero value indicate an accurate prediction. Here, the positive and negative show the overestimation. Most of the errors are very close to zero, which shows that the performance of the neural network is very good. The model reveals high accuracy and stability, as most prediction errors are minor and steadily distributed across all datasets. Regression graphs with R values of 1 for training, validation, testing, and total data are shown in Fig. 27e. These plots demonstrate the correctness and dependability of the ANN model and prove an incredible correlation between expected and actual outputs. Figure 27f contains two subplots that evaluate how well the neural network predicts the target values. This plot makes comparisons between network outputs (predictions) and the true target values over the input domain. Oscillations in errors are closely around zero, demonstrating no systematic bias present. Figure 28a–f is assessing the train ANN model of concentration profile , here, the and are input parameters and the concentration profile is out parameter. Figure 28a displays the excellent agreement between the bvp4c and ANN approximation. Figure 28b shows the best validation performance of mean square error at 6.4987 and epoch 09. Figure 28c demonstrates the validation check at epoch 09. Figure 28d shows the Error Histogram of Neural Network Performance for Training, Validation, and Test Sets. This is commonly used for performance of the Artificial Neural Network. This figure in a histogram represents the difference between the predicted output and the target (true) value. The values near the zero value indicate an accurate prediction. Most of the errors are very close to zero, which shows that the performance of the neural network is very good. The model reveals high accuracy and stability, as most prediction errors are minor and steadily distributed across all datasets. Regression designs with R values of 1 for training, validation (0.99986), testing (0.99975), test (0.99986), and total data (0.99984) are shown in Fig. 28e, supporting a significant correlation between likely and actual outputs and demonstrating the accuracy and consistency of the ANN model. ANN architecture for predicting and based on inputs M a–f. Plots for proposed Neural Network Fitting Tool (nftool) in MATLAB for solving versus a–f. Plots for proposed Neural Network Fitting Tool (nftool) in MATLAB for solving versus Figure 28f contains two subplots that evaluate how well the neural network predicts the target values. This plot makes comparisons between network outputs (predictions) and the true target values over the input domain. Oscillations in errors are closely around zero, demonstrating no systematic bias present. Comparison of results for current results validation In this section, numerical solution of proposed model with solution technique used in it are compared with published results available in the literature to observe the validation of the current solution. In Table 3 for special cases numerical results computed in current study with published are displayed. Both results are showing that there is an excellent closeness between both of them leading to the conclusion, results for current research study are numerically correct and hence have validation. This close comparison shows the accuracy of the currently employed numerical technique and algorithm in commuting software. All results asymptotically satisfy the prescribed boundary conditions. Table 3Comparison of , for a specific case Ishak et al. [44]Present0.11.28881.28870.018.49248.49130.00162.163762.1626 Comparison of , for a specific case Conclusion This study inspects the bio-convective flow behavior of a ternary Casson fluid over an inclined moving thin needle embedded in a porous medium, incorporating the effects of Arrhenius activation energy and non-uniform heat source/sink. The main outcomes are summarized in below: It was obtained that velocity field monotonically decreases when the Casson fluid flow parameter increases. It was obtained that velocity field monotonically decreases when the Casson fluid flow parameter increases. It has been found that increasing chemical reaction parameter causes velocity to drop monotonically. It has been found that increasing chemical reaction parameter causes velocity to drop monotonically. The velocity profile thereby drops as the activation energy parameter increases. The velocity profile thereby drops as the activation energy parameter increases. It is detected that the velocity cuts with growing values of thin needle. It is detected that the velocity cuts with growing values of thin needle. It was found that the velocity falls as the bio-convection Lewis number rises. It was found that the velocity falls as the bio-convection Lewis number rises. Velocity falls as Peclet number increases. Velocity falls as Peclet number increases. It has been determined that the velocity rises with cumulative values of the angle of inclination. It has been determined that the velocity rises with cumulative values of the angle of inclination. It is observed that the temperature rises as the chemical reaction parameter grows. It is observed that the temperature rises as the chemical reaction parameter grows. The temperature rises somewhat as the activation energy parameter grows. The temperature rises somewhat as the activation energy parameter grows. It was viewed that increasing the activation energy parameter leads to an enhancement in the concentration distribution within boundary layer, accompanied by a thickening of the concentration boundary layer. It was viewed that increasing the activation energy parameter leads to an enhancement in the concentration distribution within boundary layer, accompanied by a thickening of the concentration boundary layer. Bio-convection Lewis number was observed to greatly increase the spread of microorganisms close to the surface, which is followed by a slow decline across boundary layer. Bio-convection Lewis number was observed to greatly increase the spread of microorganisms close to the surface, which is followed by a slow decline across boundary layer. Increasing Peclet number is found to greatly increase the concentration of microorganisms close to the surface, which then gradually decreases throughout the boundary layer. Increasing Peclet number is found to greatly increase the concentration of microorganisms close to the surface, which then gradually decreases throughout the boundary layer. Results show influence of bio-convection Lewis number on Skin friction coefficient; it can be detected that as bio-convection Lewis number rises, the skin friction coefficient decreases for all of the cases. The maximum curve for the skin friction coefficient is obtained for a simple fluid. Results show influence of bio-convection Lewis number on Skin friction coefficient; it can be detected that as bio-convection Lewis number rises, the skin friction coefficient decreases for all of the cases. The maximum curve for the skin friction coefficient is obtained for a simple fluid. In Artificial Neural Network modelling analysis, it has been noted that the model reveals high accuracy and stability, as most prediction errors are minor and steadily distributed across all datasets. In Artificial Neural Network modelling analysis, it has been noted that the model reveals high accuracy and stability, as most prediction errors are minor and steadily distributed across all datasets. Limitations of the current study In the current research study, two dimensional, steady, incompressible and viscous laminar fluid flow is assumed. In the current research study, two dimensional, steady, incompressible and viscous laminar fluid flow is assumed. In the current work single phase of nanofluid model has been considered. In the current work single phase of nanofluid model has been considered. In the current analysis Casson fluid model has been considered. In the current analysis Casson fluid model has been considered. In current study uniform thermophysical properties are considered. In current study uniform thermophysical properties are considered. In the current work nanoparticles dispersion is uniformly assumed. In the current work nanoparticles dispersion is uniformly assumed. In this problem, uniform and perpendicular magnetic field is assumed. In this problem, uniform and perpendicular magnetic field is assumed. Inclined thin needle embedded in a porous media has been assumed in the current flow model. Inclined thin needle embedded in a porous media has been assumed in the current flow model. In current work gyrotactic microorganism density assumption has been considered. In current work gyrotactic microorganism density assumption has been considered. Here, in current problem, the non-unfirm heat source and sink are assumed. Here, in current problem, the non-unfirm heat source and sink are assumed. In present analysis the effects of shape of nanoparticles have not been accounted. In present analysis the effects of shape of nanoparticles have not been accounted. The analysis focuses primarily on theoretical and computational aspects. The analysis focuses primarily on theoretical and computational aspects. The Artificial Intelligence methodology is employed mainly for predictive approximation. The Artificial Intelligence methodology is employed mainly for predictive approximation. Future directions The current model can be extended to three dimensional and unsteady flow. The current model can be extended to three dimensional and unsteady flow. The current model can be extended to turbulent flow. The current model can be extended to turbulent flow. In future extension the shape effects of nanoparticles can be enhanced. In future extension the shape effects of nanoparticles can be enhanced. Other non-Newtonian fluid models such as Maxwell, Carreau, Williamson, Oldroyd-B, Cross, and Jeffrey fluids can be examined for comparative analysis. Other non-Newtonian fluid models such as Maxwell, Carreau, Williamson, Oldroyd-B, Cross, and Jeffrey fluids can be examined for comparative analysis. In the future extensions Darcy–Forchheimer and anisotropic porous structures, can be incorporated. In the future extensions Darcy–Forchheimer and anisotropic porous structures, can be incorporated. This work can be extended for quaternary nanofluids flow model. This work can be extended for quaternary nanofluids flow model. Variable thermophysical properties in extended model can be assumed. Variable thermophysical properties in extended model can be assumed. In the future direction, current problem can be extended for inclusion of the non-linear thermal radiation, entropy generation and inclined induced magnetic field can be incorporated. In the future direction, current problem can be extended for inclusion of the non-linear thermal radiation, entropy generation and inclined induced magnetic field can be incorporated. The shrinking and stretching sheets, rotating disks etc., can be incorporated in the extended model. The shrinking and stretching sheets, rotating disks etc., can be incorporated in the extended model. In future convective boundary conditions can be included in the model. In future convective boundary conditions can be included in the model. Acknowledgements Princess Nourah bint Abdulrahman University Researchers Supporting Project number (PNURSP2026R730), Princess Nourah bint Abdulrahman University, Riyadh, Saudi Arabia.The authors extend their appreciation to the Deanship of Scientific Research at Northern Border University, Arar, KSA, for funding this research work through the project number “NBU-FFR-2026-2505-04”. List of symbols Axial and radial coordinates Axial and radial velocity components, respectively Velocity field Stream function Fluid temperature Needle/wall reference temperature Ambient/environment temperature Species concentration in the fluid Needle/wall species concentration Ambient/environment concentration Microorganism density Wall/reference microorganism density Ambient microorganism density Fluid density Nanoparticle density Microorganism density Dynamic viscosity Kinematic viscosity Specific heat at constant pressure Thermal conductivity Electrical conductivity Thermal expansion coefficient Concentration expansion coefficient / microorganism-volume coefficient as used in the buoyancy term Porous-medium permeability/porosity parameter Applied magnetic field strength Gravitational acceleration Inclination angle of the surface/needle Volumetric heat generation/absorption term Non-uniform heat source and heat sink parameters, respectively Ratio of nanoparticle heat capacity to base-fluid heat capacity Brownian diffusion coefficient Thermophoretic diffusion coefficient Molecular diffusivity Chemical reaction rate constant Activation energy Boltzmann/Arrhenius constant used in the activation-energy term Reaction order Maximum cell swimming speed Chemotaxis constant in the microorganism conservation equation Reference velocity Needle/wall velocity Free-stream/ambient velocity Local needle radius/surface position Volume fractions of Al2O3, TiO2, and SiO2 nanoparticles, respectively Similarity variable Dimensionless stream function Dimensionless velocity profile Dimensionless temperature profile Dimensionless concentration profile Dimensionless microorganism density profile Casson fluid flow parameter Mixed convection parameter Magnetic field parameter Buoyancy ratio parameter Bioconvection Rayleigh number Darcy number Prandtl number Brownian motion parameter Thermophoresis parameter Schmidt number Bioconvection Lewis number Eckert number Peclet number Velocity ratio parameter Needle thickness parameter Temperature ratio parameter Chemical reaction parameter Dimensionless activation energy parameter Space- and temperature-dependent non-uniform heat source/sink coefficients in the transformed energy equation Skin-friction coefficient Local Nusselt number Local Sherwood number Local microorganism-density number Local Reynolds number Subscripts Wall/needle surface condition Ambient/free-stream condition Base fluid, water Mono nanofluid Hybrid nanofluid Ternary hybrid nanofluid Nanoparticle Microorganism Abbreviations Artificial neural network MATLAB boundary-value-problem solver Boundary value problem Computational fluid dynamics Magnetohydrodynamics Mean square error Ordinary differential equation Partial differential equation MATLAB neural network fitting tool Author contributions A.A., S.K.S., M.F. and S.M.: conceptualization (equal); methodology (equal); supervision (equal); writing-original draft preparation and funding acquisition (equal). A.E. and N.A.: resources and investigation (equal); writing-original draft (equal); writing-review and editing (equal). W.H. and L. K.: resources (equal); investigation (equal); figures and conceptualization (equal); writing-review and editing (equal). All authors agreed for possible publication. Funding Princess Nourah bint Abdulrahman University Researchers Supporting Project number (PNURSP2026R730), Princess Nourah bint Abdulrahman University, Riyadh, Saudi Arabia. The authors extend their appreciation to the Deanship of Scientific Research at Northern Border University, Arar, KSA, for funding this research work through the project number “NBU-FFR-2026-2505-04”. Data availability The data that supports the findings of this study are available within the article. Declarations Ethical approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Clinical trial number Not applicable. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The data that supports the findings of this study are available within the article.