Keywords

7.1 Introduction

The conventional Hydraulic Ram Pump (HRP) depicted utilises two check valves that enable the system to generate cyclic changes in pressure [1]. The water hammer generated can be manipulated to pump water to a higher elevation than its initial source in the expense of a portion of expelled water. The cycle is described by three phases: The first phase, ‘acceleration’, supplies water through the body and discharges at the waste valve. Once the flow reaches a critical velocity, the drag forces acting upon the waste valve are sufficient to close the check-valve, trapping the high dynamic pressure flow to rapidly transfer into static pressure. This hammer surge in pressure breaches the delivery valve, injecting a small volume of high pressure water into the delivery line. This pumps water against the elevation pressure, known as the ‘delivery’ phase. To restart the cycle, the ‘recoil’ phase is a resultant of the drop in pressure within the device following delivery. This reopens the waste valve and enables the flow to accelerate out of the device once more. The air chamber is implemented to manage the peaks in pressure and results in a continuous stream of pumped water [2, 3].

In recent years, our understanding and approach towards the climate crisis has escalated, with ambitious commitments set across the globe [4]. Consequently, technologies powered by renewable sources have increased in demand and research has been reinvigurated into both the conventional system and alternative Hydro Powered Pump (HPP) mechanisms [5]. The following piece of research centres around an emerging alternative design, developed recently in Australia by Brett Porta and Ralph Glockemann called the Bunyip ‘pressure amplifier’ perpetual piston pump. Significant contributions were made by the author Young in [3] and [6]. This work continues to be frequently used today to support the numerical analyses undertaken, improving our understanding of the internal flow details and phenomena within HRP. The first piece of research to discuss, aimed to understand the influence of the waste valve entry region shape upon its performance [7]. The research identified that local diffuser enlargements performed best by reducing the head loss coefficient, drag coefficient and velocity uniformity across the component. The paper additionally highlighted that the abrupt changes in flow area and direction induce vortex regions, negatively influencing the performance. Consequently, conventional designs require precisely manufactured parts to achieve higher efficiency values, increasing their cost and complexity of manufacture. In what follows, the proposed system is designed with parametric analysis, using ANSYS-Fluent CFD software tool [8].

7.2 System Design

The initial stage of modelling was to extract the fluid body within the system and appropriately defeature the design in such a way that will simplify the modelling process, whilst being conscious not to introduce significant systematic error within the model. The system is powered solely by water, developed over the last decade through incremental invention [9, 10]. One of the founders had moved to the ‘Australian Bush’, opting to construct a HRP to supply water from a local creek [11]. However, with insufficient supply head and intermittent flow, failed to meet its requirements [12, 13]. Consequently, a series of concepts were developed, soon translating into commercial success, developing several models including the Oasis, Water Dragon and the Glockemann Pump, achieving a Gold Medal at the Geneva 2002 International Exhibition [14] and [15]. The system remained to be noisy and disruptive to the local area. Thus, in an attempt to reduce the undesired characteristics, the water hammer phenomena could be removed, leading to a series of redesigns and eventually the Bunyip PA-13 [8, 16–19], depicted in Fig. 7.1.

Fig. 7.1
An illustration of a circular H P P. Its components are labeled bore, bore flange, output, flexible output hose, supply pipe rubber coupling, anchoring eye, base, vent ball valve, output ball valve, pumped water inlet filter, and pump inlet one-way valve.

Annotated Bunyip HPP [20]

Full size image

To attain a representative geometry, engagement with the pump manufacturers enabled a scaled cross section of the large Bunyip PA-13 model to be shared, depicted in Fig. 7.2. Initially, the known length could be taken from the 100mmØ supply pipe and used to determine the geometry throughout the system.

Fig. 7.2
A cross-sectional illustration of Bunyip P A-13. It consists of springs, a piston, a cylinder bore, a pump chamber, a piston rod, and a tire.

Bunyip PA-13 scaled cross-section drawing

Full size image

Following [21], when defeaturing the design, the following assumptions were made to reduce the modelling complexity, without introducing systematic errors from the real system. Initially within the lower region of the pump, the area below the inlet pipe can be assumed to be flat, without the requirement to model the fixings for the internal springs and fixed rod length, considered to have negligible influence upon the internal flow of the system [22, 23]. Additionally, the tyre mechanism deformation would add significant complexity to the wall definition and processing activities of the model, outlining the initial locations and profile of movement throughout the calculation. For this reason, the tyre has been constructed as a constant diameter cylinder that will simulate deformation and changes in volume through the extension of the cylinder height. Within the upper region, several defeatures could also be made. Firstly, the inlet and outlet check valves if modelled could introduce associated systematic error, without time to appropriately study their operation within the scope of research. Therefore, the valves will be manipulated as inlets and outlets, synchronised with the anticipated position of the valves throughout the cycle. Additionally, a series of discharge holes are located within the internal bore, highlighted within Fig. 7.2 using an asterisk, enabling the piston to discharge once it has descended beyond these points. The modelled geometry relocates the discharge point to the underside of the piston, eliminating the required interaction of the sliding piston and discharge locations, whilst ensuring the same access is available for water to freely exhaust from the volume.

7.3 Results

An execute commands have been defined and embedded within a Scheme file script of ANSYS-Fluent [8, 24]. This utilises a series of ‘IF’ logic statements and the flowtime within the calculation to manipulate TUI commands to define the boundary zone types and parameters [25]. The times for each phase could be determined, starting with a larger period than required and iteratively estimating the period using the previous result. At this stage, each of the inlets and outlets could be set, using velocity, pressure, and mass-flow methods, suited for use with incompressible water. The zone details are provided in Table 7.1.

Table 7.1 Boundary conditions for the associated system inlets and outlets
Full size table

The post-processing of the CFD-Post module is conducted within ANSYS [26], enabling the efficient production of surface contours, vectors and streamlines for the velocity and pressure within the system [27]. The content will be discussed within Sect. 4. System Velocity Contours are shown in Fig. 7.3. The lower stream is shown in Fig. 7.4, the delivery streamlines are shown in Fig. 7.5, the upper piston operation is illustrated in Figs. 7.6 and 7.7.

Fig. 7.3
6 illustrations of the contours of the Bunyip cycle velocity in different color gradients. Their respective cycle times are 1 tilde 0.09 seconds, 2 tilde 0.74 seconds, 3 tilde 0.99 seconds, 4 tilde 1.03 seconds, 5 tilde 1.19 seconds, and 6 tilde 1.34 seconds.

Bunyip cycle velocity contour illustrations with local axis, annotated [number ~ cycle time]

Full size image
Fig. 7.4
6 illustrations of the Bunyip cycle tire expansion in different color gradients. Their respective cycle times are 1 tilde 0.01 seconds, 2 tilde 0.09 seconds, 3 tilde 0.441 seconds, 4 tilde 0.89 seconds, 5 tilde 1.09 seconds, and 6 tilde 1.29 seconds.

Bunyip cycle tyre expansion lower region streamline illustration, annotated [number ~ cycle time]

Full size image
Fig. 7.5
3 illustrations of the Bunyip cycle piston streamlines in different color gradients. Their respective cycle times are 1 tilde 0.03 seconds, 2 tilde 0.09 seconds, and 3 tilde 1.01 seconds.

Bunyip cycle piston delivery, exhaust and inlet streamlines, annotated [number ~ cycle time]

Full size image
Fig. 7.6
3 illustrations of the normalized velocity vectors and contours in different color gradients. Their respective cycle times are 1 tilde 0.04 seconds, 2 tilde 0.99 seconds, and 3 tilde 1.02 seconds.

Normalised (*non) velocity vectors and contours, to depict the flow mechanics within the upper region, annotated [number ~ cycle time]

Full size image
Fig. 7.7
4 illustrations of the normalized velocity vectors and contours in different color gradients. Their respective cycle times are 1 tilde 0.01 seconds, 2 tilde 0.89 seconds, 3 tilde 1.09 seconds, and 4 tilde 1.44 seconds.

Normalised velocity vectors and contours to depict the flow mechanics within the lower region

Full size image

In order to validate the model, two data sets will be used to contrast against the outputs attained through previous stages. Firstly, the publicly available Bunyip output chart enables direct reference for the supply rate, supply elevation and delivery head of 6 L/s, 4 m and 50 m as modelled. Additionally, having contacted the manufacturer, an advanced calculator tool was shared, used to quote customers, and advise the most suitable system for application [28]. The two data sets could be summarised for validation in Table 7.2.

Table 7.2 Bunyip validation data
Full size table

7.4 Conclusions

The Bunyip dynamic motion was fully defined using the 6 degrees of freedom (DOF) solver, incorporating both the Bunyip mass and the internal spring stiffness, manipulating the mesh using the layering method to construct/destroy mesh layers. Once the model was processed the results were analysed and illustrated using the integrated CFD-post module to produce a series of figures and plots discussed. The model produced was successfully validated against two data sets, achieving agreement within the region of 10% for both the daily output and supply efficiency, recognised to consistently underestimate output parameters as a consequence of a slightly reduced piston diameter, deemed appropriate for the current research application. Despite the lack of information to validate the waste efficiency, the model emphasises that the larger diameter waste valve for pressure amplification will naturally exhaust larger volumes to provide some of the distinguishing features of the Bunyip. Consequently, it is expected that the Bunyip waste efficiency, determined at ~ 12%, is unlikely to achieve any greater than 30% waste efficiency. Thus, for water scarce applications, may not be as viable as precision made industry HRP alternatives achieving ~ 60% or more. Beyond quantitative data, the model enabled the flow mechanics and characteristics of the Bunyip to be visualised and discussed to elaborate upon operation, identifying potential ways to improve the function and share further research opportunities for the Bunyip system.