Home / microfluidic research summaries / T-Junction Microchannel flow dynamics : the influence of flow rate pulsation
Microfluidic research summary

Published on 24 April 2025

T-Junction Microchannel flow dynamics : the influence of flow rate pulsation

The present study performed in Kutateladze Institute of Thermophysics of the Siberian Branch of the Russian Academy of Sciences (Novosibirsk, Russia) describes how researchers controlled the liquid-liquid flow in their T-junction microchannel by using the OB1 Pressure Controller to apply external high-fidelity pulsations to the flow rate of a dispersed phase.

The research summary is based on the peer-reviewed article “Influence of dispersed phase flow-rate pulsations on the liquid-liquid parallel flow in a T-junction microchannel” co-written by Alexander V. Kovalev, Konstantin S. Pervunin, Artur V. Bilsky, Anna A. Yagodnitsyna and published in Chemical Engineering Journal 2024, 488, 150734 and “Control of Plug Flow Dynamics in Microfluidic T-junction Using Pulsations of Dispersed Phase Flow Rate” co-written by Alexander V. Kovalev, Anna A. Yagodnitsyna, German V. Bartkus and Artur V. Bilsky in Journal of Thermofluids 2024, V. 23, 100720. The research was funded by the Russian Science Foundation (grant No. 21-79-10307).

T-junction flow rate pulsation

Abstract

In this study, Anna A. Yagodnitsyna et al. investigated liquid-liquid parallel and plug flow under the dispersed phase flow-rate pulsations in a T-junction microchannel. To that matter, sinusoidal flow rate disturbances with variations in amplitude and period were applied to dispersed phases of different viscosities.

The study focuses on:

  • Flow visualization to observe fluid behavior in a T-junction.
  • Velocity field measurements using the micro-PIV technique under both free and excited flow conditions.
  • Comprehensive flow analysis, including flow pattern maps, velocity fields, velocity gradients, and phase-averaged velocity profiles, both in the T-junction region and further downstream.

Parallel Flow & Plug Flow Regimes

Two distinct flow regimes, parallel flow and plug flow, are investigated, both showing wave propagation disturbances influenced by the viscosity of the dispersed phase. In the parallel flow regime, the interaction between transverse and longitudinal waves varies based on viscosity of the dispersed phase, leading to different mechanisms of disturbance evolution and flow instability, governed by the Ohnesorge number. In the plug flow regime, flow rate pulsations at frequencies proportional to the natural plug formation frequency result in diverse plug length distributions, including multi-mode and drop-on-demand patterns. The impact of pulsations is particularly significant at low dimensionless frequencies (f < 1), especially for high-viscosity ratio fluids. While pulsations stretch streamlines within plugs, the plug length remains unchanged.

Overall, the study highlights how viscosity, frequency, and flow amplitude shape the dynamics of wave propagation and flow stability in both regimes.

Introduction | Changing flow rate pattern through active flow control

In two-phase microsystems, manipulating the flow to broaden or narrow the range of desired conditions is an important practical issue that urges conducting research on how various control measures, especially active ones involving external periodic disturbances, affect two-phase flow.

Segmented or plug flow in microchannels is known to be limited by the maximum flow rate of the dispersed phase or the minimum flow rate ratio of both phases. Moreover, the more viscous the dispersed phase is, the shorter the range of flow velocities, at which the segmented flow pattern exists, appears to be. This means that it is impossible to achieve a plug (segmented) flow pattern for any arbitrarily high flow rates, which limits the efficiency of microreactors operated in segmented flow regimes. Additionally, plug flow or droplet generation via passive methods fails to produce the desired plug lengths or droplet sizes across a wide range of values

In addition, the external excitation can expand the range of existence of segmented flow and increase microreactor efficiency along with accurate control over droplet breakup dynamics. Among the existing methods of active flow control, piezoelectric actuation is perhaps the most widespread and well-developed tool that, however, has a number of drawbacks. In particular, disturbances introduced by piezo elements are local in nature and depend on a specific microdevice, which does not allow the generalization of experimental data.

In contrast, herein a periodic variation of the flow rate of the dispersed phase was employed to influence the flow conditions globally. This method seems to be simpler and cheaper for implementation compared to piezoelectric actuation and, thus, represents a promising alternative for controlling the entire set of regimes.

Aim of the study

  •   Examine the effect of sinusoidal pulsations imposed on the flow rate of the dispersed phase on the stability of the parallel and the plug flow patterns of two-phase liquid-liquid flow in a T-shaped microchannel

Additional expectations: 

  •   Identify optimal frequency and amplitude of these perturbations to broaden the existence domain of stable plug flow
  •   Determine the relation between the pulsation frequency and plug or droplet size
  •   Explore the effect of the viscosity ratio of the dispersed and continuous phases on the liquid-liquid flow behavior in the T-junction microchannel under flow-rate pulsations imposed

Experimental setup | External periodic disturbances into T-junction microchannel

Figure 1a illustrates the microfluidic apparatus, the T-junction experiment was set up using a flow controller and a syringe pump to introduce the two liquids from opposite sides into the discharge duct, ensuring controlled flow conditions. It includes:

  • OB1 flow controller (Elveflow)
  • BFS flow sensor (Elveflow)
  • ESI Software (Elveflow)
  • Polymethylmethacrylate (PMMA) T-shaped microchannel (hydraulic diameter of D_h ≈ 277 µm)
  • Syringe pump
  • Vacuum pump (Labtech VP18 Plus)
  • Air compressor (Jun-Air 6-4)
T-junction flow rate figure 1
Figure 1. The scheme of experimental setup (a), a sinusoidal wave of dispersed flow rate pulsation (b).

Flow Control

  • The Elveflow OB1 pressure-based flow controller was used to drive the dispersed phase.
  • The controller generated either a constant or periodically varying pressure (up to 25 Hz), depending on the desired excitation signal (Fig. 1a).
  • A Labtech VP18 Plus vacuum pump and a Jun-Air 6-4 compressor were connected to the OB1 through flexible hoses to ensure stable pressure supply.

Flow Rate Regulation

  • A Bronkhorst BFS+ Coriolis flow meter was employed to monitor the flow rate of the dispersed phase.
  • Feedback from the flow meter was sent to the Elveflow OB1 via a computer running standard Elveflow ESI software.
  • A built-in PID algorithm automatically adjusted the pressure in real time to maintain or vary the flow rate.
  • This setup enabled precise control of flow parameters and the ability to configure custom excitation waveforms (see example in Fig. 1b).

Visual Observation of Phase Interface

  • The interface between the two liquids was made visible due to their different refractive indices.
  • A halogen lamp provided illumination, and the flow was recorded using a PCO 1200 hs high-speed camera (1.3 MP sensor).
  • The camera was mounted on an inverted Zeiss Axio Observer.Z1 microscope fitted with a 5x objective lens.

Micro-PIV Velocity Measurements

  • Fluorescent seeding particles (mean diameter: 2 µm) were added to the dispersed phase to act as flow tracers.
  • Measurements were taken in the central longitudinal section of the T-junction discharge duct.
  • The flow was illuminated using a pulsed double-cavity Nd:YAG laser.
  • Double-frame images of the tracer particles were captured with an ImperX Bobcat GEV cross-correlation camera, also mounted on the microscope.
  • This setup enabled the evaluation of flow velocities under both parallel and plug flow conditions.
Experimental setup photo elveflow t-junction flow rate
Experimental setup with Elveflow OB1 flow controller and BFS flow sensor

Key findings | Effect of flow-rate pulsation on liquid-liquid flow pattern

First, the undisturbed two-phase flow was visualized, revealing three distinct flow patterns (Fig. 2).:

  •   segmented (plug flow)
  •   transient periodic (irregular plugs of variable size),
  •   parallel flow

Based on these visual observations, flow pattern maps were constructed for various liquids used as the dispersed phase. These maps were then generalized using a universal dimensionless complexdefined as We0.4∙Oh0.6, which proved applicable across different liquid–liquid systems with varying physical properties.

T-junction figure 2
Figure 2. Characteristic patterns of the liquid-liquid flow observed in the T-junction microchannel.

Undisturbed flow

Further analysis was conducted on the parameters of stable plug flow, including plug length, velocity, and related characteristics. The behavior of undisturbed plug flow in a T-junction microchannel was compared with existing correlations for plug length, velocity, and the thickness of the continuous phase film. Within the range of tested flow rates, plug formation was observed in two regimes:

  •   the “squeezing” regime for capillary numbers of the continuous phase (Cac < 0.01)
  •   the “transient” regime for Cac > 0.01.

(NB: Cac = Capillary number, a dimensionless parameter that quantifies the balance between viscous drag forces and surface tension forces at the interface between a gas and a liquid, or between two immiscible liquids.)

Plug velocity and bulk velocity

The relationship between plug velocity and bulk velocity (defined as Ubluk = Ud + Uc) was evaluated for all fluid sets and is presented in Fig. 3a. An empirical approximation for plug velocity was found to be valid across all tested fluids:

U plug equation, flow rate

where C1 and C2 are the constants that depend on the viscosity of the continuous phase and the microchannel geometry.

Plug formation frequency

The natural frequency of plug formation ƒplug, was determined as the average number of plugs generated over a given time interval. Estimating the value of ƒplug from initial flow conditions was considered essential prior to experimental measurements. In consequence, an equation for the effective frequency of plug formation was derived based on certain simplifications. 

feff equation, flow rate

This effective frequency of plug formation is determined by the superficial velocities of the dispersed and continuous phases, Ud  and Uc, and the microchannel width, w. 

As shown in Fig. 3b, the relationship between the measured plug formation frequencyplug) and the effective formation frequencyeff) was examined. It was found that a power-law function accurately describes this relationship. Moreover, the actual formation frequency ƒplug consistently exceeds the effective frequency and increases following a power-law trend with respect to ƒeff​. This behavior is attributed to the thickening of the continuous phase film, which aligns with the observed trend in plug velocity.

T-junction figure 3
Figure 3. Plug velocity (a) and plug formation frequency (b) in an undisturbed flow (SOX are the continuous phase caracteristiques, find the details in Control of plug flow dynamics in microfluidic T-junction using pulsations of dispersed phase flow rate, A. Kovalev et. Al, 2024)

OB1 flow controller

High precision and responsiveness for your flow and pressure regulation

Stability of the disturbed parallel flow

As the control parameters were varied, the two-phase microflow in the parallel regime was observed to lose stability under the influence of imposed flow-rate pulsations. Initially, the interface exhibited wave-like deformations, which eventually led to breakup and the formation of a plug flow regime. Notably, interfacial waves decayed more rapidly when the dispersed phase consisted of a less viscous liquid.

This phenomenon was interpreted through the lens of the Ohnesorge number of the dispersed phase Ohd.

  • At higher Ohd values, short disturbance waves are efficiently damped by viscous forces.
  • For low ​Ohd liquids, the disturbances tend to interact with the interface, leading to the peeling of the dispersed phase layer from the channel wall and altering the segmentation mechanism of the parallel flow.

Consequently, a unifying dimensionless complex was proposed to generalize the observed flow patterns (Fig. 4).

It was observed that velocity gradients associated with shear stresses were significantly higher for the less viscous liquid, with differences reaching approximately an order of magnitude. As a result, viscous stresses exerted a stronger influence in cases where the dispersed phase had lower viscosity, further impacting the stability and dynamics of the two-phase flow.

T-junction figure 4
Figure 4. Unified flow pattern map of the externally disturbed two-phase microflow for high-viscous dispersed phase, where A is the non-dimensional amplitude of flow-rate pulsations, T_e is the period of the external disturbances, T_1 is the characteristic flow time, λ is the viscosity ratio of the liquids and 〖Ca〗_c is the capillary number of the continuous phase. The x-axis represents the non-dimensional capillary number (Ca).

Plug flow under flow rate pulsations  

To investigate the interaction between naturally occurring plug or droplet generation at constant flow rates and externally imposed pulsations, the plug flow pattern under pulsatile flow conditions was systematically studied. As a reference point, the natural plug formation frequency in undisturbed flow, denoted as ƒplug, was used. This allowed for the definition of a dimensionless frequency, ƒ= ƒpulseplug represents the frequency of the imposed pulsation signal, chosen to be proportional to ƒplug.

The analysis focused on key parameters expected to be affected by pulsations, namely the plug length (Lplug), plug velocity (Uplug), and their standard deviations. A general trend of increasing average plug length and its standard deviation was observed for ƒ<1, although the precise form of this dependence on the dimensionless frequency remained unclear.

 To clarify these observations, plug length distributions over time were analyzed. The evolution of plug length as a function of time is shown in Fig. 5, along with the corresponding histograms. It was found that the largest increase in standard deviation occurred in cases where double- or triple-mode distributions emerged. These distribution patterns were particularly evident at ƒ=1/2 and ƒ=1/4, although they did not appear uniformly across all amplitude values.

T-junction figure 5
Figure 5. Different regimes of plug formation under pulsations, Qc = 2 µl/min, Qd = 1 µl/min. Left – time sequences of plug length, right – histograms of plug length. Corresponding visualization results can be found in the Supplementary materials.

Considering the observed similarity between the resulting plug length distribution patterns and the phenomenon of modulation—characterized by the superposition of two harmonic signals—a mathematical model was derived to describe plug length behavior under pulsatile flow rate conditions.

Here, Tplug = 1/ƒplug is the natural period of plug formation, and φ0 represents the natural signal phase, Tpulse is a pulsation period, φ1 – the disturbance signal phase, and CA is the signal amplitude, which generally represents a function of pulsation amplitude A and liquid properties CA = CA(A, µc, µd,…). The coefficient CA was determined for all investigated regimes based on the experimental data. Representative examples demonstrating the application of Equation (1) are presented in Fig. 6.

A good agreement between model (1) and the experimental results was observed for both gas–liquid and liquid–liquid flow systems. Each identified plug length distribution could be explained by the superposition of signals with varying amplitudes. According to the model, plug detachment occurs at the local maxima of the combined signal.

The coefficient  CA was observed to increase with rising pulsation amplitude A, while other parameters were held constant. Additionally, the influence of the fluid viscosity ratio λ was noted, along with a clear dependency of CA on ƒ and the flow rate ratio, establishing these as key functional arguments of the model. By applying the proposed model, it becomes possible to qualitatively describe system behaviors such as the emergence of double-mode distributions or the drop-on-demand mode, where a single, consistent plug length is generated. These flow patterns offer potential for use in microfluidic logic systems and other applications where controlled flow rate pulsations can be employed to regulate flow characteristics.

T-junction figure 6
Figure 6. Comparison of different regimes of plug formation under pulsations with equation (1), Qc = 2 µl/min, Qd = 2 µl/min.

Micro-PIV measurement

To assess the impact of flow rate pulsations on the internal velocity fields of plugs, micro-PIV measurements were performed using a phase-averaging approach. Representative velocity distributions obtained from these measurements are presented in Fig. 7. In the central cross-section, viewed in a reference frame moving with the plug velocity Uplug the flow structure revealed two counter-rotating circulation zones, resembling the classic pattern observed in Taylor flow.

Importantly, this symmetry remained intact even under the highest pulsation amplitudes tested. However, significant stretching of streamlines was observed near the maxima of the pulsation signal. Given that plug length remained largely unaffected by pulsation parameters within the studied range for ƒ>1, it was concluded that such pulsations can be effectively used to enhance mixing and mass transfer in T-junction microchannels without altering plug dimensions dictated by the channel geometry.

T-junction figure 7
Figure 7. Velocity and vorticity fields in plugs (in the reference frame of plug) and the velocity profiles corresponding to the dashed line under flow rate pulsations in different phases of pulse signal, Qc = 4 µl/min, Qd = 2 µl/min, f = 1, A = 2.

Design your own PDMS chips

With Elveflow PDMS Chip Replication Station

Conclusions | Flow rate pulsation in T-junction microchannel  

In summary, a flow control method was proposed and implemented, based on the application of external sinusoidal perturbations to the dispersed phase flow rate. This approach enabled effective manipulation of liquid–liquid flow within a T-shaped microchannel, with both the amplitude and period of the input signal varied, while the continuous phase flow rate remained constant.

  • In the parallel flow regime, when the less viscous liquid serves as the dispersed phase, transverse waves amplify longitudinal ones due to the relaxation of the liquid-liquid interface. Conversely, in the high-viscosity dispersed phase, longitudinal waves are significantly damped, while transverse wave amplitude remains nearly constant. This distinction led to the identification of two distinct mechanisms governing disturbance wave propagation and the loss of parallel flow stability, both influenced by the Ohnesorge number of the dispersed phase.
  • In the plug flow regime, imposing flow rate pulsations at frequencies proportional to the natural plug formation frequency resulted in distinct plug length distributions, including double- and triple-mode distributions, drop-on-demand formation, and multi-mode patterns. The influence of pulsations on plug length was particularly pronounced at low dimensionless frequencies (f < 1), especially for fluid systems with higher viscosity ratios. A model was developed to explain plug length distribution patterns under pulsatile conditions. Micro-PIV measurements further revealed that while streamlines within plugs were stretched under pulsatile flow, plug length remained unchanged. 

The findings lay the groundwork for the development of active control strategies for managing two-phase microflows. The proposed dimensionless parametric complex and the plug length model under pulsatile flow offer practical tools for precisely adjusting control parameters in future research and applications.

Overall, the use of this active control strategy can expand the accessible range of segmented flow patterns, even in systems involving high-viscosity dispersed phases, thereby potentially enhancing the performance of microreactors. Furthermore, it allows for the generation of plugs with targeted lengths and can significantly enhance mixing and mass transfer—capabilities that are crucial for various chemical and biological processes. 

Elveflow instruments for T-junction studies 

The advantage of Elveflow instruments in this T-junction setup lies in their ability to precisely monitor and control flow rates on demand, whether constant or periodically varying. To achieve this, the OB1 flow controller was used for its high accuracy in pressure control, paired with the BFS+, the most precise flow meter in our range, which provides feedback to the controller to ensure stability. The ESI software enabled full control of the system, allowing the creation of custom flow rate profiles. In this setup, the dispersed phase was regulated by the OB1 flow controller, while the continuous phase was initially injected using a syringe pump. However, thanks to the autonomy of the OB1’s independent channels, the syringe pump can be replaced by a second OB1 channel to control the continuous phase in parallel. All our systems are modular and upgradable—don’t hesitate to get in touch with us!

Authors Information

t-junction team photo
German Bartkus, Konstantin Pervunin, Artur Bilsky, Anna Yagodnitsyna, and Alexander Kovalev
German Bartkus, Research Fellow at the Kutateladze Institute of Thermophysics SB RAS, received his PhD degree in fluid mechanics from Novosibirsk State University in 2023.
Konstantin Pervunin, Lead Engineer at the Kutateladze Institute of Thermophysics SB RAS, earned his MS degree in physics from Novosibirsk State University and completed his postgraduate studies at the Kutateladze Institute of Thermophysics SB RAS in 2010.
Artur Bilsky, Senior Research Fellow and Deputy Director of the Kutateladze Institute of Thermophysics SB RAS, received his PhD degree in thermophysics from the same institute in 2006.
Anna Yagodnitsyna, Senior Research Fellow at the Kutateladze Institute of Thermophysics SB RAS and Senior Lecturer at Novosibirsk State University, received her PhD degree in fluid mechanics in 2019
Alexander Kovalev, Research Fellow at the Kutateladze Institute of Thermophysics SB RAS and Teaching Assistant at Novosibirsk State University, earned his PhD degree in fluid mechanics from Novosibirsk State University in 2023.

Full publication: Influence of dispersed phase flow-rate pulsations on the liquid-liquid parallel flow in a T-junction microchannel” co-written by Alexander V. Kovalev, Konstantin S. Pervunin, Artur V. Bilsky, Anna A. Yagodnitsyna and published in Chemical Engineering Journal 2024, 488, 150734 and “Control of Plug Flow Dynamics in Microfluidic T-junction Using Pulsations of Dispersed Phase Flow Rate” co-written by Alexander V. Kovalev, Anna A. Yagodnitsyna, German V. Bartkus and Artur V. Bilsky in Journal of Thermofluids 2024, V. 23, 100720.

Written and reviewed by Anna Yagodnitsyna and Louise Fournier, PhD in Chemistry and Biology Interface. For more content about microfluidics, you can have a look here.

    For more information or Technical discussion


    Contact
    How can we help you?
    Quoteor technical request Job application Job
    application
    Collaboration or partnerships Collaborations
    or partnerships
    Customer support Customer
    support
    Others questions Other

      For more information or Technical discussion


      Contacting for
      a job application?
      We are happy that you are interested in Elveflow. You can apply to our open jobs or send us your open application on WelcomeToTheJungle. Over here!

        For more information or Technical discussion


          Need customer support?







          I hereby agree that Elveflow uses my personal data

          [recaptcha]



            For more information or Technical discussion