Abstract

An equilibrium stream is characterized as one where channel dimensions and slope have been adjusted over a duration, enabling the conveyance of incoming sediment and water with minimal erosion or deposition. The time taken by the flow to attain this equilibrium state is referred to as equilibrium time. Studies have been conducted on cohesive channel bed for critical shear stress expression and further extended for bed load transport rate; however, studies were not found on equilibrium time expression on erosion for cohesive channel bed. The present study fills the gap for the computation of equilibrium time for cohesive mixture of clay–silt–sand. A laboratory-based investigation was conducted to analyse the factors influencing the time it takes for equilibrium to be reached in a cohesive channel bed made of mixture of clay, silt, and sand in which clay content varied from 10% to 50%. The study indicated a decreasing trend in effective shear stress with an increase in clay content. An increase in clay content in the channel bed led to a lengthening of the equilibrium time. A relationship was established to compute the dimensionless equilibrium time for the cohesive clay–silt–sand mixture. Impressively, a strong correlation (R² = 0.97) emerged between the computed values and the actual observed data in this study. The developed model estimates the dimensionless equilibrium time in cohesive clay–silt–sand mixtures, emphasizing the dominant influence of clay content on erosion dynamics and bed stabilization.

1 Introduction

The present study deals with the model development based on statistical analysis using experiential data in respect of space and time for the management of sediment on the surface of a channel bed. The movement of sediment commences from the mobile channel bed once the shear stress generated by the incoming flow surpasses the critical shear stress of sediment on that channel bed. Transport of sediment increases, in general, as the shear stress by incoming flow increases from the critical shear stress of the sediment. It has been noted that in the beginning, the sediment transport rate is elevated (Singh and Ahmad 2019). Nevertheless, this rate diminishes over time, ultimately reaching a point where minimal to no sediment transport is evident. This marks the channel bed's attainment of an equilibrium state. An equilibrium stream is characterized as one where channel dimensions and slope have been adjusted over a duration, enabling the conveyance of incoming sediment and water with minimal erosion or deposition (Mackin 1948). Garde and Ranga Raju (2000) provided an alternate definition for equilibrium within a stretch, stating that it exists when the sediment entering the stretch is equal to the sediment exiting it. However, any deviation from this equilibrium state leads to changes in the alluvial river system's channel shape, potentially resulting in aggradation or degradation. These alterations persist until the equilibrium condition is established, reinstating the balance between incoming and outgoing water and sediment flows. The time taken by the flow to attain this equilibrium state is referred to as equilibrium time. Equilibrium time is pivotal in sediment transport within open channel flow. It denotes the duration necessary for a watercourse to achieve a balance among sediment erosion, transportation, and deposition. Establishing equilibrium time is imperative for understanding sediment transport dynamics in open channels. Equilibrium time is directly linked to the stability of the channel bed and signifies the timeframe required for the channel to stabilize its morphology under sediment transport influences.

Sediment transport within alluvial channels bears a significant impact on the longevity of hydraulic structures situated within these channels (Pandey et al. 2017). Alluvial rivers naturally undergo ongoing adjustments, involving changes in cross-sectional shape, bed elevation, slope, and more, in response to shifts in sediment input. Human activities expedite these changes in river courses. Instances of such alterations due to human intervention include sediment deposition upstream of dams, the lowering of channel bed downstream of dams, localized scour around hydraulic structures, among others. The volume of sediment carried by the flow in an alluvial channel profoundly influences the erosion of the channel bed within alluvial rivers. Minimal or virtually absent sediment within the incoming flow can lead to "hungry water" or sediment-deprived water, potentially resulting in channel incision that poses a risk of undermining bridges and other constructions. As noted by Garde and Ranga Raju (2000), the degradation of Islam weir on the Sutlej River is attributed to degradation phenomenon. Similarly, in the Yellow River, a reduction in bed level by an average of 4.5 m across a 50-km stretch downstream of the dam occurred due to the presence of less heavily sediment-laden water (Garde and Ranga Raju 2000).

Channel bed degradation may result in the destruction of aquatic and riparian habitat by causing bed coarsening (Eddy et al. 2007). Moreover, it is a threat to water structures such as bridge piers as it causes local scour very close to them (Pandey et al. 2018). Hekal (2018) evaluated the dynamic morphological equilibrium status of the River Nile in terms of riverbed and plan form. Ahmed and Fahmy (2014) compared the morphological changes before and after the construction of Aswan High Dam along the reach of the Nile River and concluded that bed aggradation occurred after the construction of the dam. Sediment transport within a channel can have detrimental effects, such as shortening the anticipated lifespan of hydraulic structures. Kothyari (1996) reported a bridge, over a torrent crossing the Dehradun-Mussoorie Road in Uttarakhand, lost around 70% clearance below its soffit due to aggradation.

The accumulation of sediment has the potential to obstruct navigation channels, leading to potential flooding. In situations where fine sediment settles within gravel spaces, it obstructs water flow between the gravel, a phenomenon known to reduce oxygen levels critical for benthic organisms, as highlighted by Owens et al. (2005). Julien (2002) documented the creation of an armour layer as an indicative marker of stable bed conditions in a channel with a cohesionless bed, particularly when the flowing water in the channel is devoid of sediment input. The degradation of the bed is a response to the stream's carrying capacity until a stable bed state is achieved. In this state, characterized by the armour layer, sediment transport is non-existent, aligning with the equilibrium condition. The pace of channel bed degradation diminishes over time, a phenomenon observed by Vogel et al. (1992), implying a gradual increase in total load with time. Until equilibrium is attained, sediment transport rates may fluctuate considerably. Thus, knowledge of equilibrium time enhances predictions of sediment transport behaviour.

Equilibrium time also sheds light on erosion and deposition patterns along the channel bed, facilitating the management of sediment-related concerns. Moreover, it serves as a fundamental component in sediment budgeting, enabling the comprehensive assessment of sediment sources, transport pathways, and deposition sites within a river system. Duan and Xu (2023) highlighted the importance of the study of the stability of the middle reaches of the Huaihe River in China to the river’s regulation, planning, and flood control. They predicted the evolution trend of that riverbed. Doyle and Harbor (2003) studied the effect of the presence of sediment type on the equilibrium condition of the reach in the channel bed. They found that transport of sediment is more rapid in sand-bed channels compared to the gravel-bed channels and concluded that quicker equilibrium conditions were present for sand-bed channels. This indicates that the equilibrium time varied as per the presence of the type of sediment, especially for cohesionless and cohesive sediment. The transport behaviour of cohesive sediment is significantly different from that of cohesionless sediment in the channel. Mitchener and Torfs (1996) noted that the existence of mud within the range of 3% to 15% prompts a shift in erosion behaviour from cohesionless to cohesive. In a study by Aberle et al. (2004), it was observed that erosion rate follows an exponential decline over time. They concluded that this erosion rate variation hinges on bed material characteristics including dry bulk density, water content, clay content, and sand content. The alterations in the clay percentage within the sediment mixture also impact the degradation of the channel bed (Singh and Ahmad 2019). In a study by Jain and Kothyari (2009), an investigation was carried out concerning the transport of a cohesive sediment mixture consisting of clay-gravel and clay-sand-gravel. Jain proposed an equation for bed load transport, wherein sediment input into the channel was absent and sediment transportation took place owing to the excess shear stress generated on the channel bed due to the in-flow of clear water. Studies were conducted on cohesive channel beds for critical shear stress expression by various researchers (Jain and Kothyari 2009; Ahmad et al. 2018. The studies were further extended for bed load transport rate (Jain and Kothyari 2009; Singh and Ahmad 2019), however, there were limited studies found for equilibrium time on erosion. Singh and Ahmad (2019) studied bed profile computation that included the equilibrium time for the clay-silt-gravel mixture. Singh et al. (2023) proposed a formulation to compute equilibrium bed load for cohesive mixture of clay–silt–sand. One of the most significant problems for river channels is the lack of accuracy in sediment transport equations (Gomez and Church 1989). The objective of the present study is to fill the gap by proposing the expression for the computation of equilibrium time for cohesive mixture of clay–silt–sand.

2 Experimental Set-Up and Procedure

The tilting flume used in the study had dimensions of 16 m in length, 0.75 m in width, and 0.50 m in depth. This experiment took place at the Hydraulic Engineering Laboratory within the Civil Engineering Department of the Indian Institute of Technology Roorkee, situated in Roorkee, India. Within the flume, a specific test section measuring 6 m in length, 0.75 m in width, and 0.18 m in depth was established. This test section commenced 7 m from the flume entrance. The depth of this test section, where a cohesive bed of clay–silt–sand mixture was introduced, was maintained at 0.18 m. To replicate the roughness of this sediment-filled test section, a thin and even layer of sediment was spread across the rest of the flume bed. Water flow within the flume was controlled by a valve positioned in the inlet pipe, which was connected to an overhead tank. A rectangular tank was placed downstream of the flow deflector to gauge the discharge. Additionally, a tail gate was positioned at the end of the flume to manage the flow depth. To assess the suspended load, a depth-integrated sampler was installed just before this tail gate. The channel's side wall was constructed using glass, enabling observation. For bed level measurements, a two-dimensional bed profiler installed over the flume's railing. An HR Wallingford-manufactured two-dimensional (2D) bed level profiler with accuracy and resolution of ±0.5 mm in both vertical and horizontal directions was employed to assess the channel's bed configuration. This profiler system is comprised of various components, namely a support beam, a carriage for the profiler, a probing device, a power supply unit, and a computer, illustrated in Figure 1. The support beam is affixed above the specific bed area where the bed profile measurements were conducted. Operating along the length of the flume, the profiler carriage moves along the support beam. A vertical mobile probe is situated at the front of the profiler carriage. In response to provided instructions, the probe descends at the designated location, contacting the bed to capture a measurement.

Figure 1  2D bed profiler installed over flume railing.

For the water surface profiles, a pointer gauge with a precision of 0.1 mm is used to measure the water surface profile. Measurements for both bed and water surface profiles were taken at intervals of 0.5 m along the flume's centre line. During the initial stages of the run, when the degradation of the bed level was rapid, profiles were captured at 15-minute intervals. Subsequently, as the bed transients became more gradual, measurements were recorded at intervals of 30, 60, and 120 minutes. During the initial stage of each run, sediment transport rates and bed degradation were high, with rapid changes in bed profile. Therefore, measurements were taken at 15-minute intervals to adequately capture this transient phase. With the progress of time, the rate of bed change decreased and approached equilibrium conditions. Accordingly, the sampling interval was increased (30–120 minutes), as bed evolution became gradual and required less frequent observation. In this study, cohesive sediment mixtures of clay–silt–sand were used, the clay component's weight percentage was systematically varied from 10% to 50%, while maintaining equal proportions of the cohesionless sediments. The range of arithmetic mean size of clay, silt, and sand mixture are reported in Table 1, while particle sediment distribution follows Singh et al. (2023). The channel bed in the test section was prepared by taking the dry weight of sediment as per proportions and then manually mixed with water. After mixing, the sediment was covered with polythene and left for 24 h for uniform moisture distribution and finally filled in the test section and compacted in three equal layers. Each layer was compacted by passing a cylindrical roller over it with a weight of 400 N, while the sides of channel were compacted by a hand rammer with a rectangular bottom (Ahmad et al. 2018). Bed preparation has been done under control conditions by allowing the equal number of passes (ten passes, i.e., five rounds) of cylindrical roller over the test section in each layer. To enhance bonding between layers, the top surface was roughened with a trowel before the subsequent layer was applied. Following the compaction of all three layers, excess sediment was removed using a large, sharp-edged knife. The prepared cohesive bed was then allowed to settle for approximately 16 h to facilitate cohesive bonding between the cohesive and non-cohesive matrix. Samples were extracted from the downstream section of the cohesive bed to determine their bulk density, unconfined compressive strength, and moisture content. Before the beginning of the experimental run, bed was saturated for 24 h to achieve the field’s condition. The range of experimental data tabulated in Table 1 for the cohesive mixture of clay–silt–sand.

Table 1  Range of data on equilibrium time for cohesive mixture of clay–silt–sand.

3 Results and Discussion

The commencement of erosion from the channel bed occurs once a specific flow rate or discharge is reached within the flume. Initially, the degradation from the channel bed is more rapid and subsequently diminishes over time, attributed to the observed decrease in bed load accumulation in the trap installed just after the tail gate in the flume. At the start of the experiment, the bed is relatively undisturbed. The initial high rate of bed degradation is attributed to the presence of excess shear stress acting on a relatively unstructured surface. As erosion progresses, the bed undergoes restructuring and partial armouring, leading to increased resistance due to cohesive bonding. Concurrently, the effective shear stress reduces, resulting in a gradual decline in the rate of degradation and sediment transport over time. The sampling intervals were selected to capture the full temporal evolution of bed adjustment without introducing bias in the estimation of equilibrium time. A point is eventually reached, typically after a certain duration, where fluctuations in the bed level degradation become insignificant, as depicted in Figure 2. Additionally, both the transportation of bed load and suspended load substantially decrease, as illustrated in Figure 3. Equilibrium time is defined based on the stabilization of the bed profile and a negligible temporal change; the denser early sampling combined with extended later observations improves its accuracy for equilibrium time. The state is identified as the equilibrium condition, characterized by minimal bed level variation and a sediment transport rate that has dwindled to less than 2% of the initial rate (Singh and Ahmad 2019). In the present study, equilibrium time is primarily defined based on the stabilization of bed profile and a negligible temporal change, rather than the 2% threshold of initial transport rate. The 2% threshold of the initial sediment transport rate is used merely as an indication to support this condition, rather than as the sole defining criterion. It was observed that channel bed stabilizes when the transport rate, in general, is between 1–2% of the initial transport rate. Further, it was observed that a less than 1% threshold leads to overestimation of equilibrium time, as it delays the identification of channel bed stabilization. In contrast, thresholds of 5% and 10% result in underestimation, as they identify equilibrium prematurely while measurable bed adjustments are still ongoing. The period required to attain the equilibrium state is defined as the equilibrium time. In the context of the present study, the equilibrium time varied in relation to the clay percentage within the sediment mixture. Notably, it was observed that an increase in clay percentage led to an elongated equilibrium time, as highlighted in Figure 4.

Figure 2  Bed degradation along the channel for clay–silt–sand (SSC) mixture.

Figure 3  Temporal variation of sediment transport rate for clay–silt–sand mixture with 50% clay content at Q = 0.03 m³/s, S₀ = 0.00272.

Figure 4  Variation of equilibrium time with clay percentage for clay–silt–sand mixture.

The following parameters have been considered in the development of formulation for the computation of equilibrium time (Equation 1).

(1)

Where:

Equation 1 can be represented in the dimensionless form using dimensional analysis as seen in Equation 2, below:

(2)

Where:

=

dimensionless equilibrium time and computed using Equation 3:

(3)

Where:

=

dimensionless excess shear stress on the test section of channel bed and computed as Equation 4:

(4)

Equation 2 delineates the dimensionless form of the functional relationship governing equilibrium time. Equation 3 indicates the equilibrium time in dimensionless form, while Equation 4 indicates the excess shear stress in dimensionless form. The exponent 1/7 applied to  in Equation 3, arises from a scaling consideration. The magnitude of  in the dataset is of the order of 10^{^}7, and taking the 1/7 power, effectively normalizes it to an order of one or zero. This transformation reduces data spread, improves numerical stability, and enhances regression performance without altering the underlying physical relationships. The parameter is incorporated to account for the diverse clay percentages' effects, while the parameter accommodates the sediment transport dynamics. The trends of parameters concerning the dimensionless equilibrium time are illustrated in Figures 5–7 for the cohesive sediment mixture of clay–silt–sand. It is important to note that the data from this study was exclusively utilized due to the unavailability of equilibrium time computational data from other researchers.

Figure 5  Variation of dimensionless equilibrium time with clay fraction.

Figure 6  Variation of  with  for different clay percentage in clay–silt–sand mixture.

Figure 7  Comparison between computed and observed dimensionless equilibrium time.

The data presented in Figure 5 demonstrates that the dimensionless equilibrium time extends as the clay percentage escalates for the clay–silt–sand sediment mixture investigated in this study. The presence of increased clay content in the cohesive channel bed introduces a heightened resistance against erosion, potentially prolonging the time required to attain equilibrium conditions.

Illustrated in Figure 6 is the consistent downward trajectory of the dimensionless excess shear stress, evident across each clay percentage. Elevating the dimensionless excess shear stress on the channel bed generates greater erosion rates, consequently expediting the attainment of the equilibrium state. An increase in clay content from 10% to 50% results in a systematic rise in bed resistance, as indicated by the upward shift of the curves as shown in Figure 6. This reflects the enhanced role of cohesive forces in strengthening inter-particle bonding. This transition defines the equilibrium condition, with higher clay content mixtures requiring longer durations to stabilize due to greater cohesive strength and highlight the critical role of clay content in controlling erosion dynamics and equilibrium time.

Following an extensive array of trials, the subsequent relationship proposed for computing equilibrium time within the context of the cohesive sediment mixture of clay–silt–sand is calculated using Equation 5:

(5)

Equation 5 outlines the formulations for calculating the cohesive sediment mixture of clay–silt–sand. The equation is derived from the data set of the current study. The dimensionless excess shear stress () is retained due to its fundamental role in sediment entrainment, particularly for cohesive beds. Although the fitted exponent (−0.0065) is nearly zero, the excess shear stress in this study varies over a relatively narrow range (0.49–0.90 SI), which limits its statistical influence in the regression.  is preserved to maintain physical consistency and ensure applicability of the model under a wider range of hydraulic conditions where its influence may become significant. The values computed using Equation 5 exhibit a strong concurrence with both the observed data and noteworthy regression coefficients, as visually depicted in Figure 7. Singh and Ahmad (2019) worked on a clay-silt-gravel mixture and the equilibrium time reported for this cohesive mixture ranges between 695–960 minutes, while the present study works on a clay–silt–sand mixture and the equilibrium time for this mixture ranges between 625–745 min. The equilibrium time is comparatively less for the clay–silt–sand mixture in the present study compared to the clay–silt–gravel mixture by Singh and Ahmad (2019). The equilibrium time is less for the clay–silt–sand mixture because the presence of sand particles increases the erodibility of the bed, allowing rapid adjustment under flowing water. In contrast, the clay–silt–gravel mixture exhibits higher resistance due to the presence of coarse gravel particles, which require higher shear stress for movement and lead to the armouring of the bed, thereby increasing the equilibrium time. The experimental study was conducted under two discharge conditions and two channel bed slopes for each clay content (10%, 20%, 30%, 40%, and 50%), with observations recorded at different time intervals, yielding a total of 320 data points. The empirical relationship (Equation 5) was developed using a nonlinear power-law regression, implemented through logarithmic transformation and ordinary least squares (OLS) fitting. The model performance is characterized by a coefficient of determination of R² = 0.97, Root mean square error (RMSE) = 0.035, and a Nash-Sutcliffe Efficiency (NSE) = 0.967, indicating a strong agreement between computed and observed values. In Figure 7, the observed and computed values closely aligned with the line of agreement (LOA) and within ±5% error lines, also indicating a strong agreement between observed and computed values.

4 Conclusion

A laboratory-based investigation was conducted to analyze the factors influencing the time it takes for equilibrium to be reached in a cohesive channel bed made of mixture of clay, silt, and sand, in which clay content varied from 10% to 50%. A formula was proposed to calculate this equilibrium time by considering the entire duration from the start of sediment transport until the sediment transport rate drops below 2% of the initial rate, as defined by Singh and Ahmad (2019). The study revealed that initially, the sediment transport rate was higher, but it gradually decreased over time. Notably, degradation was more pronounced in the upstream working section compared to the downstream counterpart. Moreover, an increase in clay content in the channel bed led to a lengthening of the equilibrium time. The study also indicated a decreasing trend in effective shear stress with an increment in clay content. A relationship was established to compute the dimensionless equilibrium time for the cohesive clay–silt–sand mixture. Impressively, a strong correlation (R² = 0.97) emerged between the computed values and the actual observed data in this study. This study is limited to a clay–silt–sand mixture in which clay content varied from 10–50%. Although the study was conducted under controlled laboratory conditions, the governing physical processes of flow and sediment transport remained consistent with field conditions. While exact field replication is not intended, the results provide process-based insights that are transferable to real-world engineering applications through appropriate interpretation. Scale effects may influence sediment transport behaviour due to differences in flow depth and particle interaction compared to natural rivers. Future research should focus on validating the experimental findings with field data to enhance their applicability to rivers. Field validation is therefore recommended as future scope of work.

References

1. Aberle, J., V. Nikora, and R. Walters. 2004. “Effects of bed material properties on cohesive sediment erosion.” Marine Geology 207 (1–4): 83–93.
2. Ahmed, F.A. and W.A. Fahmy. 2014. “Long-term morphological changes in the Nile river since high Aswan dam construction to year 2010.” Nile Water Science Engineering Journal 7 (1): 44–57.
3. Ahmad, Z., U.K. Singh, and A. Kumar. 2018. “Incipient motion for gravel particles in clay-silt-gravel cohesive mixtures.” Journal of Soils and Sediments 18, 3041–3043. https://doi.org/10.1007/s11368-017-1869-z
4. Doyle, M.W. and J.M. Harbor. 2003. “A scaling approximation of equilibrium timescales for sand-bed and gravel-bed rivers responding to base-level lowering.” Geomorphology 54 (3–4): 217–223.
5. Duan, Y. and G. Xu. 2023. “Analysis of River Stability in the Middle Reaches of Huaihe River Based on Non-equilibrium Thermodynamicsins.” In: Proceedings of PIANC Smart Rivers 2022, Lecture Notes in Civil Engineering, Vol. 264. Springer Singapore. https://doi.org/10.1007/978-981-19-6138-0_91
6. Eddy, Y., R.J. Zengb, and C. Wanga. 2007. “Effects of in-channel sand excavation on the hydrology of the Pearl River Delta.” China Journal of Hydrology 343, 230–239.
7. Garde, R.J. and K.G. Ranga Raju. 2000. Mechanics of Sediment Transportation and Alluvial Stream Problems. New Age International, New Delhi, India.
8. Gomez, B. and M. Church. 1989. “An assessment of bed load sediment transport formulae for gravel bed rivers.” Water Resources Research 25 (6): 1161–1186. https://doi.org/10.1029/WR025i006p01161
9. Hekal, N. 2018. “Evaluation of the equilibrium of the River Nile morphological changes throughout Assuit-Delta Barrages reach.” Water Science 32 (2): 230–240. https://doi.org/10.1016/j.wsj.2018.09.001
10. Jain, R.K. and U.C. Kothyari. 2009. “Cohesion influences on erosion and bed load transport.” Water Resources Research 45 (W06410): 1–17.
11. Julien, P.Y. 2002. “River Mechanics.” Cambridge University Press, New York, 434.
12. Kothyari, U.C. 1996. “Erosion and sedimentation problems in India.” IAHS Publ. 236 (Jul): 531–540.
13. Mackin, J.H. 1948. “Concept of the graded river.” Bulletin of the Geological Society of America 59: 463–511.
14. Mitchener, H. and H. Torfs. 1996. “Erosion of mud/sand mixtures.” Coastal Engineering 29 (1–2): 1–25.
15. Owens, P.N., R.J. Batalla, A.J. Collins, B. Gomez, D.M. Hicks, A.J. Horowitz, G.M. Kondolf, et al. 2005. “Fine-grained sediment in river systems: environmental significance and management issues.” River Research and Applications 21, 693–717.
16. Pandey, M., P.K. Sharma, Z. Ahmad, and U.K. Singh. 2017. “Evaluation of existing equations for temporal scour depth around circular bridge piers.” Environmental Fluid Mechanics 17, 981–995. https://doi.org/10.1007/s10652-017-9529-9
17. Pandey, M., P.K. Sharma, Z. Ahmad, and U.K. Singh. 2018. “Experimental investigation of clear-water temporal scour variation around bridge pier in gravel.” Environmental Fluid Mechanics 18, 871–890. https://doi.org/10.1007/s10652-017-9570-8
18. Singh, U.K. and Z. Ahmad. 2019. “Transport rate and bed profile computations for clay–silt–gravel mixture.” Environmental Earth Sciences 78, 432. https://doi.org/10.1007/s12665-019-8419-5
19. Singh, U.K., S. Kumar, Z. Ahmad, A. Chakravarti, S. Bhave, and M. Pandey. 2023. “Equation Development for Equilibrium Bed Load.” In: Pandey, M., Azamathulla, H., Pu, J.H. (eds) River Dynamics and Flood Hazards. Disaster Resilience and Green Growth. Springer, Singapore. https://doi.org/10.1007/978-981-19-7100-6_16
20. Vogel, K.R., A.V. Niekerk, R.L. Slingerland, and J.S. Bridge. 1992. “Routing of Heterogeneous Sediments over Movable Bed: Model verification.” Journal of Hydraulic Engineering 118 (2): 263–279. https://api.semanticscholar.org/CorpusID:140603400