Numerical Investigation of Maneuverability in Barge Transportation Using Towing and Pushing Operations
Article information
Abstract
Barge transportation using tug–barge systems is widely employed for large cargo transport in rivers and harbor areas, where maneuverability in confined waterways is critical. This study numerically investigates the maneuvering performance of tug–barge systems under pushing and towing operations. Simulations were conducted using a maneuvering model based on the Maneuvering Mathematical Modeling Group framework with experimentally derived hydrodynamic coefficients for each operation mode. For pushing operations, barge configurations were varied by changing the numbers of barges in the longitudinal and transverse directions, whereas for towing operations, the effects of skeg attachment and towline length were examined through turning simulations. The results show that increasing the number of barges generally degrades turning performance, and longitudinal extension causes a greater increase in turning radius than transverse extension. In towing, skeg attachment reduces turning performance by increasing yaw damping and restoring moment, whereas longer towlines enlarge the turning radius by producing a greater negative yaw moment on the tug. Overall, towing yields smaller drift angles and greater rudder effectiveness than pushing, indicating superior turning performance in confined waterways.
1. Introduction
Barge transportation offers the advantage of transporting large volumes of cargo in a single operation, making it widely used in confined waterways, such as rivers and harbors. Since barges typically lack self-propulsion and rely on external propulsion, the coupling method between the barge and tug directly affects the stability and maneuverability of transport. The two most common operational modes of barges are towing and pushing. In towing operations, a tug pulls a barge via a towline. While this method allows transportation without requiring an additional secure connection system (such as a notch), it has drawbacks, including instability caused by tension fluctuations and inferior maneuvering performance. Conversely, pushing involves a tug propelling the barge from behind as a single unit, which is reported to provide faster response times and superior maneuverability because the thrust is directly transferred to the barge (Sano & Hasegawa, 2015). However, it requires a separate coupling system capable of withstanding substantial loads exerted on the connection interface. Given the recent upward trend in barge traffic volumes and collision incidents, there is a pressing need for studies that systematically investigate the differences in maneuvering performance and stability dictated by these transportation modes.
Regarding towing stability, various experimental and numerical simulation studies have been conducted, primarily focusing on nonlinear behavioral characteristics, the effects of skegs and towlines, and the occurrence of slewing motion. Among notable recent studies, Yasukawa et al. (2006) established a towing simulation based on the Maneuvering Modeling Group (MMG) model, compared the results with experimental data, and evaluated towing stability. Subsequently, Yasukawa et al. (2008) conducted experiments and numerical simulations under wind-load conditions to compare changes in slewing motion according to wind direction and speed, presence of a skeg, and towline length. Hong et al. (2013) investigated the towing stability in waves. Nam et al. (2013, 2014b) compared the MMG model with a cross-flow model; quantitatively examined the effects of varying hydrodynamic coefficients, towing speed, and towline length on slewing motion; and verified their findings through experiments. Nam et al. (2014a) identified and compared the variations in the towing stability and towline tension when multiple tugs towed a barge. Nam et al. (2015) derived linear towing stability criteria under wind load conditions and utilized them to characterize changes in towing stability with respect to wind speed and direction. Furthermore, Nam (2020) identified the nonlinearity of the slewing motion across a broader range of conditions by varying the resistance and tug oscillation in addition to primary towing parameters, such as towline length, towing speed, and hydrodynamic coefficients, and by comparatively analyzing the phase trajectory of the towed vessel. Nam et al. (2022) performed numerical simulations and experiments under disturbance conditions, such as wind and waves, to analyze the variations in towing stability and phase trajectories. Research has also been conducted on barge-turning characteristics and towline dynamics during towing. Fitriadhy and Yasukawa (2010, 2011) analyzed turning performance during barge towing using MMG-based maneuvering equations of motion, while Fitriadhy et al. (2015) derived a linearized equation for the angle between the tug and towline to predict slack towline occurrences, validating this through experiments. Additionally, although Sano and Hasegawa (2015) compared the turning performances of towing and pushing, they did not structurally analyze the differences in maneuverability depending on transport conditions, such as towline length or skeg usage. Baso et al. (2021) experimentally measured the heave and pitch amplitudes of vessels towed in waves. Furthermore, Baso (2022) experimentally verified the changes in towing stability based on the damage conditions of the barge.
However, research on pushing operations has primarily focused on the effects of varying operational parameters, such as barge configuration, loading conditions, and water depth, on maneuvering and resistance performance. King et al. (2008b) analyzed changes in maneuverability based on the number and arrangement of barges during pushing operations using an MMG-based model, and Koh et al. (2008) confirmed that maneuverability varies significantly depending on the arrangement, even when the number of barges remains the same. King et al. (2008a) investigated changes in the turning performance of pushing operations according to water depth. Regarding the loading conditions, Sano et al. (2017) applied computational fluid dynamics (CFD) to calculate the resistance and connection interface flow based on the barge loading conditions in a pushing system coupling a pusher and barges. Subsequently, Sano et al. (2018) conducted circular motion tests (CMT) under various loading conditions. Based on these results, turning, zigzag, and stopping tests were performed using MMG-based numerical simulations to identify the characteristics of the maneuverability changes according to the loading conditions. In studies examining water depth effects in pushing operations, Koh and Yasukawa (2012) conducted rotating arm tests (RAT) on a pushing system at three different water depths and performed numerical simulations to comparatively analyze turning performance. Maimun et al. (2011) compared the hydrodynamic coefficients calculated from planar motion mechanism (PMM) tests with the results from empirical formulas. They verified these findings through MMG-based simulations and compared them with experimental data. Bravo et al. (2018) evaluated response amplitude operators (RAOs) and connection loads according to the wave direction through a frequency-domain analysis of the pushing system. Delefortrie et al. (2022) conducted model tests by varying the number of barges and loading conditions to measure 6-degree-of-freedom forces. Wang and Sano (2025) assessed the seakeeping performance of a multi-barge pushing system via regular wave experiments, obtained the RAOs for the motion of each barge and tug, as well as the connection loads, and compared the effects of increasing the number of barges.
In this study, numerical simulations were conducted to directly compare the maneuverability of the towing and pushing transportation modes previously examined individually and to understand their fundamental characteristics. For pushing, simulations were performed using the model test data, thrust, and rudder force coefficients obtained by King et al. (2008b), whereas for towing, the model test data reported by Fitriadhy et al. (2015) were used. Through numerical simulations, the turning performance of towing with and without a skeg was comparatively analyzed, along with the turning performance and physical forces of both the towing and pushing operations. Ultimately, this study aimed to clarify the effects of skeg attachment during barge transport, evaluate the maneuvering performance of each mode, and identify the underlying physical forces and quantities driving these characteristics.
2. Methodology
2.1 Pushing Model
In this study, maneuvering equations of motion based on the Maneuvering Modeling Group (MMG) model were applied to simulate barge behavior during pushing operations (King et al., 2008b). As it was assumed that the tug and barge were completely fixed during pushing, the integrated tug–barge system was regarded as a single floating body. Accordingly, a single-body-fixed coordinate system was utilized, as shown in Fig. 1, and the equations of motion for the coupled system can be expressed as those for a single floating body, as shown in Eq. (1):
where m and Iz represent the mass and yaw moments of inertia, respectively, of the entire integrated system. The terms mx, my, Jz correspond to the added mass coefficients in the surge, sway, and yaw directions, while u, v, r denote the surge velocity, sway velocity, and yaw rate at the center of gravity, respectively. The hydrodynamic forces acting on the hull, XH,YH,NH are expressed by Eq. (2):
In these equations, ρ is the fluid density, and L, d are the length and draft of the integrated system.
In pushing operations, barges of various widths and lengths are used, or multiple barges are operated in a connected configuration. To examine these effects, this study analyzed the influence of barge arrangement on turning performance by considering the nine barge configurations proposed by Koh et al. (2008), as shown in Fig. 2, and increasing the number of barges by up to three in both the transverse and longitudinal directions. According to the arrangement, the hydrodynamic derivatives were identical to those presented by Koh et al. (2008).
The thrust and rudder models used in the simulation are the same as those employed by Koh et al. (2008), as shown in Eqs. (3) and 4, respectively. Because this study considered a propulsion system with twin screws and twin rudders, the propeller and rudder forces were expressed as the resultant force of thrust from the two propellers and the forces from the left and right rudders. The port and starboard sides are denoted by subscripts p and s, respectively, and the moment generated by the y-axis distance is not considered. In addition, the thrust and rudder force coefficients for the port and starboard sides were assumed to be identical.
where Dp is the propeller diameter, n the number of propeller revolutions per second (rps), and tp the thrust deduction factor. KT is the thrust coefficient measured in the propeller open-water tests and is expressed by Eq. (5):
In this context, JP is the propeller advance ratio, and p denotes the propeller pitch ratio, expressed as follows:
wP is the wake coefficient determined using Hirano’s formula, as shown in Eq. (7):
Where wP0 is the propeller wake at the initial advance speed, and
δ is the rudder angle, tR is a deduction factor adjusting the effective transfer rate of longitudinal force generated by the rudder, and aH is a coefficient representing the additional lateral force due to rudder–hull interaction, xH is the moment arm of this additional lateral force. xR denotes the moment arm of lateral rudder force YR . The rudder normal force FN is expressed by Eq. (8):
In this context, AR is the rudder area,
2.2 Towing Model
In towing operations, because the tug and barge move as independent rigid bodies, separate body-fixed coordinate systems are utilized for each hull, and the equations of motion are described independently (Fitriadhy et al., 2015). The equations of motion for the tug (i = 1) and barge (i = 2) are shown in Eqs. (11) and (12), respectively. The inertial characteristics of each hull are represented by hull mass mi and yaw moment of inertia Izi, where mxi, myi, Jzi correspond to the added masses in the surge, sway, and yaw directions of the hull, respectively. ui, vi, ri correspond to the surge velocity, sway velocity, and yaw rate at the center of gravity, respectively. at the center of gravity, respectively. XHi, YHi, NHi represent the hydrodynamic forces acting on each hull; while XP is the thrust from the propeller; XR, YR, NR are the rudder forces; and XTi, YTi, NTi represent the forces and moments caused by the tension acting at the towline connection points. At this time, thrust and rudder forces exist in the case of the tug (i = 1) but do not exist for the barge (i = 2). Therefore, whereas the equations of motion for the tug include thrust and rudder terms in addition to hull hydrodynamic forces and towline tension, only hydrodynamic forces and tension exist in the equations of motion for the barge.
The hydrodynamic forces acting on the hulls of the tug (i = 1) and the barge (i = 2) are expressed by Eq. (13).
Three types of vessels–tugs, barges without skegs, and barges with skegs - were used in the simulation, and their hydrodynamic derivatives were identical to those used by Fitriadhy et al. (2013). The skegs were attached to both ends of the barge stern, with a length of 6.1 m, a maximum height of 1.5 m, and a thickness of 0.1 m.
The thrust and rudder models used in the simulation are identical to those used by Fitriadhy et al. (2015), as shown in Eqs. (14) and (15), respectively.
where Dp is the propeller diameter, n is the propeller revolution, and tp is the thrust deduction thrust deduction factor. KT is the thrust coefficient measured in the propeller open-water tests and is expressed by Eq. (16). The remaining equations were identical to those used for pushing in Section 2.1; however, the thrust and rudder force coefficients used in the simulation were adopted from the values presented by Fitriadhy et al. (2015). Furthermore, for a direct comparison between the pushing and towing in Section 3.2, a comparison was performed after unifying the thrust and rudder models with those used in the towing model.
Fig. 3 shows a diagram representing the coordinate system of the towing system. The tug (i = 1) and barge (i = 2) have fixed body-fixed coordinate systems (xi, yi), and the towline connects the tug-side connection point (xtow, ytow) to the barge-side connection point (xtowed, ytowed). The heading angle of each hull is defined as φi, and the towline angle is denoted by θ. Additionally, ltug and lbarge denote the distances between the towline connection point and the center of gravity for each vessel, and the tension in the towline direction acts as a force component in the body-fixed coordinate system of each hull. In this study, the towline tension was simulated using a simple spring model. For the spring constant k, a value of 125 kN/m was used, which was converted to full–scale from the model–scale value of 50 N/m used by Nam et al. (2013), and for the damping coefficient b, a value of 25 kN·s/m was used at the same ratio. If the instantaneous length of the towline is LT and the initial towline length is Ltowline, then the tension T is expressed as in Eq. (17), considering only tension during elongation. In this case, the instantaneous towline length LT is defined as the distance between connection points (xtowed, ytowed) and (xtow, ytow) with each hull, as shown in Eq. (18).
The towline force components and moments XTi, YTi, NTi acting on the tug and barge can be expressed as shown in Eq. (19) and Eq. (20), respectively:
Here, σ, γ represent the relative angles of the towline with respect to the heading angles of the tug and the barge, respectively, and signify the directional angles of the towline as viewed from each vessel’s body–fixed coordinate system. As shown in Fig. 2, the directional angle of the towline is defined as σ = φ1 −θ for the tug and γ = θ−φ2 for the barge.
2.3 Target Vessels
The target vessels utilized in this study were the tug and barge models commonly used by Koh et al. (2008) and Fitriadhy et al. (2015). In this case, the barge is a catamaran-type barge consisting of two barges connected in a transverse direction. For the pushing operations, the system was operated in a configuration in which the catamaran barge and tug were connected in the longitudinal direction. Table 1 lists the dimensions of the target vessels.
3. Results and Discussion
3.1 Pushing Simulation
To verify the reliability of the numerical pushing simulation established in this study, a direct comparison was performed with the simulation results presented by King et al. (2008). The test condition is a case where one tug is coupled with one barge in a pushing mode, and a 20° turning test and a 10°/10° zigzag test were performed while maintaining a revolution of 5 rps and a pitch ratio of 0.612 to achieve an advance speed of 7 knots (3.6 m/s) without a rudder angle. For comparison, the trajectories of the turning test and the time series of the heading angle and rudder angle of the zigzag test were confirmed, and maneuvering performance indices such as advance, tactical diameter, transfer, 1st overshoot angle, and response time were quantitatively compared. Fig. 4(a) shows the turning trajectory of the pushing simulation, and Fig. 4(b) shows the time-series graphs of the heading and rudder angles in the zigzag test. In Fig. 4(a), the blue solid line represents the current simulation results and the red dotted line represents the results presented by King et al. (2008b). In the turning test, a circular trajectory corresponding to a steady turning diameter of approximately 500 m was observed; in the zigzag test, the 1st overshoot angle was approximately 3.5° and the 2nd overshoot angle was approximately 5°. Fig. 5 presents a bar graph that quantitatively compares the results of the turning and zigzag tests. The turning test characteristics of the pushing simulation, such as advance, tactical diameter, and transfer, showed small errors of approximately 1.9–3.3% compared with the results of King et al. (2008b). In the zigzag test, small errors of 2.8% and 1.3% were observed for the 1st overshoot angle and response time, respectively. Overall, the results of the two simulations exhibited a high degree of agreement. Additionally, to maintain an advance speed of 7 knots (3.6 m/s) for each configuration, this study assumed that the propeller revolution speed n was fixed, and the pitch ratio of the controllable pitch propeller (CPP) was increased, as in the study by Koh et al. (2008). Table 2 lists the pitch ratios used based on the barge arrangement.
Figs. 6 (a)–(c) illustrate the turning motion trajectories over time for a single barge (11BP), a case where the number of barges increased in the transverse direction (31BP), and a case where it increased in the longitudinal direction (13BP). The simulation conditions included an advance speed of 7 knots (3.6 m/s) and a rudder angle of 20°. In the case of the single barge shown in Fig. 6(a), it is confirmed that it reaches the vicinity of 400 m in the x-direction, which corresponds to the advance distance, within 3 min. In the cases of 31BP and 13BP, in which the number of barges increased to three, the advance distance and time to reach the advance distance increased. It can be confirmed that 31BP, corresponding to Fig. 6(b), reached the vicinity of 600 m in the x-direction, corresponding to the advance distance, at approximately 4 min, whereas 13BP in Fig. 6(c) did not reach the position corresponding to the advance distance even after 5 min.
Fig. 7 (a) and (b) compare the nondimensionalized turning motion trajectories when the number of barges is increased in the transverse and longitudinal directions, respectively, in a pushing situation. To fairly compare the increase in the length and width of the vessel, nondimensionalization was performed using the length of a unit barge rather than the total length of each configuration. Overall, as the number of barges increased, the scale of the turning trajectory tended to increase; in particular, the expansion of the turning trajectory appeared larger in the longitudinal direction (Fig. 7(b)) compared to the transverse direction (Fig. 7(a)). When the number of barges was increased to two and three in the transverse direction, the steady turning radius increased by 30.7% and 90.6%, respectively, whereas in the case of a longitudinal increase, it increased by 52.8% and 157.1%, respectively. That is, the longitudinal increase showed a turning performance deterioration approximately 1.7 times greater than the transverse increase. Figs. 8 (a) and (b) show bar graphs comparing the advance and tactical diameters of the nine barge arrangements presented in Fig. 5, respectively. Both turning characteristics showed a tendency to increase as the number of barges increased. In particular, it was found that the longitudinal increase had a greater effect on turning performance than the transverse increase. When the number of barges in the longitudinal direction increased three-fold from 11BP, which had one barge in the transverse direction, to 13BP, the number of barges increased by two, and the advance distance and tactical diameters increased by 124.2% and 127.5%, respectively. Conversely, when the number of barges in the longitudinal direction increased three-fold from 31BP, which had three barges in the transverse direction, to 33BP, the number of barges increased by six, but the growth rates were 34.88% and 39.6%, respectively. As a result, the advance distance and tactical diameter of 33BP, which transports nine barges, were 6.78% and 11.2% smaller than those of 13BP, which transports three barges, respectively, indicating superior turning performance. To clearly analyze the cause of this, the hydrodynamic forces acting on the hull and rudder were compared.
Table 3 lists the changes in the hydrodynamic derivatives in the transverse and yaw directions according to the change in the barge arrangement configuration. The analysis confirmed that the main factors contributing to turning performance differed depending on the arrangement direction. First, the linear transverse damping force
Table 4 shows a comparison of the steady-state values of the nondimensionalized longitudinal and transverse hydrodynamic forces, rudder normal force, drift angle, and effective rudder inflow angle for the cases where the number of barges was increased to three in the transverse direction (31BP) and three in the longitudinal direction (13BP) with the results of a single barge (11BP). At this time, forces were nondimensionalized by
3.2 Towing Simulation
The towing simulation was validated using the turning test results presented by Fitriadhy et al. (2015). Figs. 9(a) and (b) show the turning motion trajectory and time series of the tension under the conditions of an advance speed of 7 knots (3.6 m/s), a rudder angle of 15°, and a towline length of LT = 3Lpp. At this time, the propeller revolutions to maintain the advance speed corresponded to 5 rps, and the pitch ratio was 0.4601. Fig. 9(a) compares the turning motion trajectories, where the dotted line represents the trajectory of the tug, and the solid line represents the trajectory of the barge. In this case, the steady turning radius of the tug was approximately 310 m, and that of the barge was approximately 280 m, confirming that the tug turning with its own thrust and rudder force exhibited a larger turning radius compared to the barge turning solely by tension. This is because, as shown in Fig. 3, the barge is located on the starboard side of the tug during starboard turning. This is because as the towline becomes longer, the angle σ between the tug and the towline increases and the transverse distance between the tug and the barge increases, causing the barge to turn inside the turning trajectory of the tug. As a result of analyzing the errors of the two numerical simulations, in terms of the turning trajectory, the advance distance, which is highly influenced by the initial behavior, showed a larger error than the tactical diameter, although the overall errors were very small at 3.3% and 1.3% for the advance distance and tactical diameters, respectively. Fig. 9(b) shows the time series of tension, and it can be confirmed that at the beginning of turning, increases and decreases in tension are clearly observed according to the transient behavior of the two floating bodies, which then gradually stabilize upon entering the steady turning stage. The tension peaks due to transient behavior occur at two time points: the first peak occurs when the tug initially starts from a stationary state and momentarily pulls the towline, and the second peak occurs near 120 s as the separation distance between the tug and barge widens at the point where the tug begins turning, resulting in maximum tension. These characteristics can also be confirmed in the turning trajectory shown in Fig. 9(a), where the transient interval corresponding to 50–150 s represents the section that initially deviates from the steady turning trajectory. In the tension response acting on the towline in the two numerical simulations, while Fitriadhy et al. (2015) showed a maximum tension of 13.6 t at 105.5 seconds, this study yielded a (a) Turning motion trajectory (b) Time histories of tension maximum tension of 14.74 t at 121.8 seconds, corresponding to errors of 15.5% in response time and 8.4% in the maximum value. Errors in this initial behavior are considered to stem from the difference between the lumped-mass model utilized by Fitriadhy et al. (2015) and the simple spring-based towline model of this study. However, the tension in the steady turning state was 8.73 t for Fitriadhy et al. (2015) and 8.63 t in this study, showing an error of approximately 1.1%, indicating that the simulation in this study demonstrates sufficient accuracy in an engineering sense.
Validation of the towing simulation against the results of Fitriadhy et al. (2015)
In the case of towing, the turning characteristics varied significantly depending on the towline length and the presence of skegs. Therefore, this study aimed to examine the turning performance characteristics according to the towline length and the presence of skegs, and a comparison was performed alongside the turning test results using pushing. Figs. 10(a) and (b) show graphs comparing the nondimensionalized turning motion trajectories of towing and pushing according to the presence of skegs and the towline length. Nondimensionalization was performed using the length of the barge. The simulation time was 1000 s, the advance speed was 7 knots (3.6 m/s), and the rudder angle was 15°. The barge-turning performance was compared when the towline length was set to two and three times the barge length, and according to the presence of skegs. In Figs. 10(a) and (b), it can be confirmed that the barge-turning radius increases with towing as the tow line length increases. In the comparison based on the presence of skegs, the turning radius in Fig. 10(b) without skegs is larger than that in Fig. 10(a) with skegs. Furthermore, it can be confirmed that the barge turning radius tends to increase or decrease in the same trend as the turning radius of the tug. To quantitatively verify this trend, Figs. 11(a) and (b) present graphs comparing the nondimensionalized steady turning radius and transfer. For towing, only the steady turning radius and transfer of the barge were plotted. With skegs, the steady turning radius was 20.9% and 6.7% smaller compared with pushing at towline lengths of 2Lpp and 3Lpp, respectively. Without skegs, the turning radius was 4.5% smaller than (a) Towing with skegs (b) Towing without skegs pushing when the towline length was 2Lpp, but it increased significantly to 21.7% when the towline was lengthened to 3Lpp. That is, it can be confirmed that the attachment of skegs improves turning performance, whereas increasing the towline length degrades turning performance. Meanwhile, in terms of transfer, a trend opposite to that of the steady turning radius was observed. While transfer without skegs appeared 57.3% and 59.2% smaller than pushing with skegs, it only decreased by 38.9% and 32.1%, respectively, showing a larger transfer than without skegs. This is thought to be because skegs increase the damping and restoring forces in the yaw direction to improve towing stability, but simultaneously suppress heading angle changes in the initial behavior, resulting in a larger transfer. Additionally, the transfer in towing was observed to be consistently smaller than that in pushing, regardless of the presence of skegs or the towline length. This is because, unlike pushing, where the tug and barge are combined, resulting in a large yaw moment of inertia, in towing, only the tug begins turning, leading to a faster initial heading angle development. As the tug heading angle increases, the direction of tension applied to the barge changes, increasing the transverse tension on the barge. This results in a large yaw moment being applied to the barge, and its heading angle rapidly increases, leading to a small transfer.
To analyze turning performance variations according to the presence of skegs, Table 5 presents a comparison of the steady-state values of hydrodynamic derivatives in the transverse and yaw directions, the nondimensionalized yaw moment, and γ according to the presence of skegs. The yaw moment was nondimensionalized by
The turning characteristic variations according to the towline length can be confirmed by the change in the angle between the tug and the towline. Fig. 13(a) is a time series graph comparing γ and σ according to towline length. When the towline length increased from 2Lpp to 3Lpp, the variation in the angle γ between the barge and the towline was small at less than 1°, whereas the angle σ between the tug and the towline changed significantly from −31.57° to −35.37°. That is, as the towline length increases, the magnitude of σ increases, leading to an increase in the transverse tension component and the yaw moment acting on the tug. In this case, because the moment applied by the tension to the tug is in the negative direction, the yaw rate of the tug decreases, resulting in a larger turning radius. Fig. 13(b) shows a time-series graph plotting the heading angles of the barge and tug with pushing. Up to approximately 40 s, as barge turning does not occur, it can be confirmed that the tug heading angle increases identically, regardless of the towline length. Subsequently, as the barge begins turning, differences according to the towline length appear, with the heading angles of the tug and barge at the shorter length of 2 Lpp increasing more rapidly. As previously explained, when the towline length increases, the tug yaw rate decreases, leading to differences in the tug heading angle. It can be confirmed that in pushing, the heading (a) Towing angle γ, σ (b) Heading angle φ angle increases upon starting the turn, but at a slower rate compared to towing. Figs. 14(a) and (b) show snapshots of the steady turning situations for the pushing and towing modes, respectively. As can be seen in Fig. 14(a), the heading angle develops significantly in pushing, and accordingly, the drift angle β also becomes large. Conversely, in the towing shown in Fig. 14(b), the heading-angle variation is restricted, maintaining a relatively small drift angle. This difference stems from the difference in yaw moments between the two transport modes. When pushing, the moment arm between the rudder point of application and the center of gravity was long (63.24 m), causing a large rudder moment. In contrast, in towing, the distance between the rudder point of application and the center of gravity is short (22.23 m), and, as shown in Fig. 14(b), the moment due to the towline tension acts in a direction that cancels out the rudder moment. Consequently, in towing, the development of the heading angle is suppressed and the drift angle is maintained small, and this difference in the drift angle acts as a major factor determining the difference in turning performance between the two methods.
Table 6 compares the drift angle, effective rudder inflow angle, and nondimensionalized transverse rudder force for the tug during towing and pushing. Transverse rudder force was nondimensionalized by
4. Conclusion
In this study, numerical simulations using an MMG-based maneuvering model were performed for pushing and towing, two representative operation modes for transport barges, and the turning performance and physical mechanisms in each transport mode were systematically reviewed. To this end, changes in the hydrodynamic forces, drift angles, and towline angles according to major design variables such as the presence of skegs, towline length, and barge arrangement configuration were analyzed.
In the pushing mode, changes in thrust emerged as the primary cause of the difference in turning performance. When the number of barges increased in the transverse direction, the hull resistance increased significantly, which led to an increase in the thrust required to maintain the same speed, resulting in an increase in the rudder force owing to the increased rudder inflow velocity. In the case of 31BP, which represents a transverse increase, a pitch ratio approximately 30% larger than that of 13BP, a longitudinal increase, was required. Because of this increase in the control force, the transverse increase showed a turning performance approximately 1.7 times superior to the longitudinal increase.
In the towing mode, the presence of skegs and towline length emerged as the main factors determining the turning performance. The skeg is an appendage that improves towing stability by increasing yaw damping and restoring forces; in turning situations, it improves both turning performance and advance speed by keeping the angle (γ) between the barge heading and the towline small, thereby increasing the longitudinal component of tension. Conversely, as the towline length increased, the negative tension moment acting on the tug increased, which decreased the yaw angular velocity and consequently increased the turning radius.
As a result of the direct comparison between pushing and towing, towing tended to exhibit a relatively superior turning performance by lowering the drift angle by up to 73% compared with pushing, as the tension moment offsets the rudder moment, thereby increasing the rudder force efficiency by more than twofold. Furthermore, owing to the interference of the barge during the initial behavior, the transfer of towing was always smaller than that of pushing.
In this study, the effects of barge transport mode, arrangement, and presence of appendages on turning performance were quantitatively compared, and their physical causes were analyzed. These results are expected to serve as useful basic data for establishing design and operational strategies for barge transport systems operating in confined waterways. Future research requires a high-accuracy towline dynamics analysis that can more realistically simulate towline behavior, and there are plans to perform additional numerical simulations for complex transport scenarios where towing and pushing are performed simultaneously, considering actual operation conditions.
Notes
Bo Woo Nam served as an editorial board member of the Journal of Ocean Engineering and Technology. However, he had no role in the decision to publish this article. No potential conflict of interest relevant to this article was reported.
This research was supported by a Korea Agency for Infrastructure Technology Advancement (KAIA) grant funded by the Ministry of Land, Infrastructure, and Transport (Grant RS-2023-00250727) through the Korea Floating Infrastructure Research Center at Seoul National University. This work was also supported by the Korea Institute for Advancement of Technology (KIAT) grant funded by the Korean Government (MOTIR) (RS-2023-KI002688, HRD Program for Industrial Innovation). This research was part of a project titled ‘Fostering Talent in Advanced Ship Blue Tech (RS-2025-02221147)’, funded by the Ministry of Oceans and Fisheries, Korea.
