3D Printing in Hydraulic-Machinery Research

By Joel Sundström, Postdoctoral researcher, Luleå University of Technology, Sweden

In a previous blog post, it was discussed how AFC4Hydro and its participants benefit from computational fluid dynamics simulations (CFD). In this blog post, we will discuss another tool that AFC4Hydro benefits from; namely, 3D printing.

Fig 1 - 3D CAD figure of the test facility at LTU

At LTU, studies of vortex breakdown and the RVR has been ongoing since 2015 through different small student projects. An existing turbine rig for an undergraduate course in which students performed turbine performance tests was modified to investigate the RVR, and potential methods to mitigate it. During the initial experimental studies, no RVR could be found in the draft tube with the original design, and independent of turbine operation. It was not until a group of students who performed their master’s thesis conducted CFD simulations that it was realized that the water passage between the guide vanes and runner blades of the existing rig was ill-designed. The water passage together with a new runner were re-designed by means of CFD and finally 3D printed. With this new configuration, an RVR experiencing both rotating and plunging modes was observed in the draft tube, highlighting the powerful combination of CFD and 3D printing. 

To support the development of different mitigating strategies within AFC4Hydro, a test rig has been developed at LTU. The new test rig will allow a better repeatability of the measurements and detailed investigations of the flow.  Most of its part have been designed using numerical methods, and selected parts will be 3D printed, including a flow straightener and the turbine runner. Being able to 3D print the runner presents a major advantage over standard milling methods because of the complexity of the blades shape.  In addition, the cost will be significantly lower, particularly taking into consideration that multiple runners, with different blade angles, will be printed to mimic an adjustable blade turbine.  Other parts such as the draft tube cone have also be 3D printed to decrease time between numerical design and experimental testing.

Model testing has played a crucial role in the development of hydropower during the last 100 years. These tests provide detailed information about turbine performance, cavitation properties etc. that can be hard to obtain at a full-scale machine; however, these tests are associated with a high cost. Optimizing a turbine runner configuration, for example, can also be a time-consuming task if at each design step one has to send the runner for manufacturing. On the other hand, if one has access to 3D printing technology, a new design can be realized rapidly with the auxiliary benefit of being significantly cheaper.

Fig 2 -Velocity diagram figure

3D printing and teaching

3D printing is also useful when it comes to teach students about hydropower. The way that Francis turbines are typically introduced involves a rather abstract two dimensional drawing of velocity triangles shown in the left figure below. Students usually comment this figure like “I don’t understand the figure”. Being able to show students a 3D printed Francis turbine as seen on the right figure below, and from that explain the flow through the turbine considerably simplifies the exposition and thus works as an indispensable asset to increase their understanding.

Fig 3 - 3D printed turbine


About the AFC4Hydro video

By Morten Kjeldsen, Managing Director Flow Design Bureau AS, Norway.

The AFC4Hydro video addresses some of the motivations for the continued development of novel solutions for the hydropower business and therefore of the reasons for the existence of EU sponsored projects such as AFC4Hydro. 

In the video certain assertations are made and this blog-post follows up by providing the necessary references and introduction to the number crunching.

Fig 1- The AFC4Hydro introductory video. About the video: Editing was provided by Capix, Stavanger, Norway and Phil Durnford provided the voice-over. The video of Tyssedal hydropower museum was made available by Dag Endre Opedal@Kraftmuseet.

On energy production

The main reference used is the “Energy Statistical Pocketbook 2019” published by the European Commision. Data from “Statistics Norway” adds the contribution of Norway. The electricity production in Europe is annually about 3300 TWh, or 3300 billion kWh, while Norway produces about 147 TWh annually. Renewables constitute about 1000 TWh (Europe) and 144TWh (Norway), and the share of hydropower is about 330 TWh (Europe) and 138 TWh (Norway). As will be explained below hydropower can have added value in addition to being a source of energy production.

On driving cars on renewable energy

It seems to be a consensus that electrical vehicles require on average about 15kWh per 100km. From this point some mathematics can be applied. If all hydropower electricity in Europe is used to propel cars the expression would be (330 + 138) billion kWh annually divided by 15 kWh per 100km. The result becomes 31 billion times 100km annually, or 3100 billion km per year. The latter number corresponds to about 715 round-trips between the orbits of Earth and Neptune! The number of cars in Europe is about 280 millions, and when dividing 3200 billion km by 280 million the result shows 11100 km per car every year i.e. more than the 10000 km cited in the video.

On need for flexibility

Nordpool (https://www.nordpoolgroup.com/) offers services related to power trade but also keeps historical and available statistics related to power trades within and between countries. Using their web-based service, the number for consumed and produced energy per hour at a given week for Norway is downloaded and displayed, see the Graphics below. Although wind is starting to have an impact in Norway, most of the electricity production is by hydropower. As can be seen in the Graphics, a variation during any given day exists. Isolated for Norway a match between produced and consumed power is rarely found. This discrepancy is due to import/export of electricity with neighboring countries. The production in Norway, for the period shown, is at about 70% (about 35GW total installed Hydropower capacity) and changes in production are in the tenths of percentage points over a few hours. This limited data set illustrates that power generation needs to adapt to variation in consumption, to non-variable power sources such as wind and solar feeding the grid, but also to disruptions, such as trips for larger thermal or nuclear power units. Reservoir based hydropower, with its relatively rapid ramping of power setting, plays a key role in compensating for short term discrepancies between produced and consumed power.  

Fig 2- Production and consumption of electricity per hour for a week in February 2020. Data from Nordpoolgroup.com

On cost of flexibility

As stated in the previous section the hydroturbine units have a high flexibility although with a cost. By nature, hydroturbines are designed to have an optimum performance in a limited power range around the best efficiency point. Once outside of this range the efficiency drops and deleterious effect such as cavitation and excessive pressure can appear. Below a graph tries to summarize these effects and consequences. The AFC4Hydro, and similar, technologies try to rectify this by mitigating deleterious effects and improve efficiencies for all power settings. This way AFC4Hydro, and similar projects, contribute to continue development and improvement of hydropower as a beneficial source of energy and power in the European energy mix. The project aims at technologies that can be retrofitted into existing units, but similar technologies will likely be included with new unit deliveries.

Fig 3 - A busy graph that shows acceptable and unacceptable cavitation and levels of pressure pulsation and corresponding allowable operating range, and efficiencies before AFC4Hydro type technologies. The operating range and efficiencies are expected to improve when the technologies are applied.

On AFC4Hydro sites

One minutes into the video a drone is flying over Sørfjorden and shows the Tyssedal hydropower museum. The entrance to the Oksla powerplant, where AFC4Hydro technologies will be validated, is very close to this building. In addition, activities on actual turbine units will take place in Porjus and Älkvarleby.

Fig 4 - Tyssedal hydropower museum


Fluid Structure Interaction

By Rafel Roig Bauzà , PhD student, Universitat Politècnica de Cataluny

Fluid Structure Interacion (FSI) is a multiphysics problem to be solved when a structure vibrates within a fluid. For that, the coupling of the laws describing the structural mechanics and the fluid dynamics must be considered. This phenomenon is relevant when interactions between a deformable or moving structure and an internal or surrounding fluid flow are affecting both fields.

What is FSI used for?

FSI can be used, for instance, to study the structural deformation of a solid body (stresses and strains) when it is submerged in a flow (velocities and pressures) and vice versa.
FSI is present in most engineering fields:

  • Aeronautics
  •  Hydraulic machinery
  •  Biofluids
  • Offshore structures
  •  Piping systems

Fig 1- CAD image of the fluid (transparent) and Structural (brown) domain

Fig 2- Structural mode shap

What are the most common approaches?

Historically, the main approaches used to study FSI phenomena have been experimental and theoretical. However, the availability of economical and powerful simulation tools and equipment has given engineers and researchers the opportunity to approach this problem using numerical tools.
Here below we briefly present the different numerical methodologies based on the level of complexity to approach FSI problems:

• Coupled Acoustic Structural FEM
In this approach small but fast structural deformations (vibrations) lead to pressure waves in the fluid medium which is considered to be still. In this case, the coupling is bidirectional since the pressure waves also determine the body vibrations.


• One-Way FSI
This approach is used when small and slow structural deformations do not affect the fluid flow. So, the hydraulic loads induced by the flow on the body are simulated assuming a rigid boundary and then applied to the structure to determine its deformation.

• Two-Way FSI
This approach is used when large structural deformations lead to significant changes on the fluid pressure and velocity fields. Therefore, the coupling is bidirectional which means that both the fluid flow affects the structure deformation and the structure deformation affects the fluid flow.

Fig 3- Francis Turbine Runner

Role of FSI in AFC4Hydro

As it has been explained, FSI is present in most engineering problems and the AFC4Hydro project outcome is not an exception. 

To understand the dynamic response of a hydraulic turbine under different operation conditions, AFC4Hydro relies on both numerical and experimental FSI analysis. In particular, One-Way and Coupled Acoustic Structural FEM numerical approaches will be used.


A bouquet of reasons for applying CFD

By Morten Kjeldsen, Managing Director Flow Design Bureau AS, Norway.

AFC4Hydro relies on numerical analysis and Computational Fluid Dynamics (CFD) both for increasing our understanding of the root problem and for designing optimal solutions that mitigate the same problems. As for AFC4Hydro partners, CFD has a strong foothold in research and engineering groups both within academia and industry dealing with flow systems. This blog-entry discusses reasons for using and when to use CFD.

WhY cfd?

1 The first reason given is Evaluation and optimization of designs and typically at an early stage when developing solutions. For a CFD engineer this is the dream scenario and tools for parametrization of geometry and design and optimizing software can be employed. A pitfall at this stage can be the omission of secondary flows that has a profound impact on the performance of the final and realized system, and incorrect boundary conditions when evaluating performance. The alternative to CFD at this stage is usually experimental investigations. Historically the experimental approach has been considered expensive and time consuming, but with the advent of 3D printing the model scale physical testing might be revived and become a competitor but also a companion to CFD works.

2 The next reason for doing CFD is to Substantiate and visualize the solution. Results from 3D CFD is highly visual, and to the point that some claims the “C” in CFD is an abbreviation for colorful! But it is a fact that this visualization is powerful and assists both in understanding and for communication. The visualization can be used for engaging the end client, to achieve acceptance for a selected concept and even for pure marketing purposes.

Fig 1- CFD visualization of flow through a Y- junction in a hydropower plant.

3 Once a design or concept is settled, and even physically exists, other reasons for CFD emerge. One typical task is Documentation of performance. In most commercial deliveries the equipment is expected to meet a minimum performance for typical operating conditions of the whole system. Instead of creating a rigid test regime CFD can be used to extrapolate existing data. These results can then be used for a complete documentation of performance and provide input for control solutions. In the extreme case CFD can be used for Verification of conditions not possible to achieve during e.g. factory acceptance test. A typical task would be to calculate performance for conditions exceeding the range used during design.

4  The final reason given is Troubleshooting of system performance.  Often system performance failure occurs in interfaces between different products and sub-systems, although the degraded performance is experienced mostly in one specific sub-system or component. Usually the supplier of the component experiencing the trouble is given the task of investigation, but this is also the time when an independent party can enter the stage. Troubleshooting, although highly unwelcomed, usually makes the CFD engineer feel more alive! Creativity is needed to establish the root cause of the problem, but also for suggesting solutions to rectify the problem.

Fig 2- Initial, and pre- AFC4Hydro, CFD studies for design of the IPM (WP1) concept. Calculation by Prof. Cervantes at LTU.

An industrial trend is to use CFD as part of the running operation and monitoring of flow systems. This trend materializes in terminology such as digital twin, reduced-order-models and edge-computing, and is a trend characterized by the blend of sensor input and on-site numerical simulation.

AFC4Hydro is present in most stages for the use of CFD, although we try to avoid a trouble shooting scenario!

 The final test of AFC4Hydro technologies will include the use of digital twins as part of the active flow control (AFC). On the other hand, the main tasks of AFC4Hydro is more about steel, manufacturing, installation in prototype hydro-turbines and on-site presence and testing. Still these tasks are expected to proceed without too many obstacles, and mostly due to comprehensive CFD efforts in the early stages of our H2020 project.


AFC4Hydro partners contribute to the digitalization wave in hydropower

By Morten Kjeldsen, Managing Director Flow Design Bureau AS, Norway.

AFC is short for Active Flow Control and implies that information for given states of the hydroturbine operation trigger flow control systems that have a positive impact on instantaneous turbine performance. This process requires sensor input, data acquisition and analysis and a control algorithm. In the bigger picture these process steps coincide with that associated digitalization. All partners of AFC4Hydro contribute to the digitalization of hydropower. This blog entry highlights the digitalization efforts of the partner Flow Design Bureau AS (FDB).

In hydropower plants the typical measurements focus on fiscal and on protective parameters. The former includes reservoir levels and power-production by the generator, while the latter includes vibration or shaft dynamics, and temperature of critical components such as bearings and power-electronics. If protective measures exceed pre-set thresholds the system shuts down. One emerging trend involves the gathering of data for condition monitoring. At this stage a few approaches present themselves: Use of data already available from protective measurement, an Internet of Things approach by adding a high number of data-sources, and finally measurements that allow high fidelity acquisition of specific states of the machinery or system. The latter type measurements also take part in a digital twin philosophy. Another emerging measurement trend challenges existing limitations on operation of the hydro-turbine. Such limitations can relate to cavitation and vibration conditions. In practice the enforcement of these limits base itself on pre-defined permissible load ranges. The continuous measurement of cavitation intensity and shaft dynamics allows to operate the unit on actual unit feed-back and not within a constrained pre-defined range. By constantly challenging load range limitations hydropower continues to be an attractive and a flexible power supplier to the grid.

Most employees at FDB have their engineering degrees doing experimental flow studies. One consequence being an instinctive approach to management of measurement chains consisting of sensors, data acquisition, analysis and presentation of result. FDB also delivers commercial measurements for the hydropower industry. One such measurement is the thermodynamic efficiency measurements. This is an approved and standardized method that also provides an accurate determination of flow rate through the unit. The term thermodynamic refers to the fact that the measurement relies on measuring the temperature increase due to the inefficiency of the unit. The temperature increase measures in the tens of milli-Kelvin, but when performed correctly allows the determination of unit efficiency with error-bands of the order of +/- 1% for each operation point measured.

Fig 1- Accurate measurements in an industrial environment. Sensors meet harsh environment giving demands to the sensor itself, mounting and cable routing. FDB assists with all aspects, such as sensor-mounting and cable works, before data acquisition and further processing of signals.

The application of FDB’s knowledge and experience to the current and emerging needs of the Hydropower industry creates an environment for the development of beneficial solutions, services and products. A joint development project with Statkraft, another partner in AFC4Hydro, resulted in solutions such as instruments for cavitation intensity and continuous and real-time efficiency. The same project developed the framework for distributed acquisition of time-resolved data, edge processing and data transfer. The edge processing on the continuous data stream will as a minimum provide statistical data for each sampled channel/ sensor, controlled down-sampling of time-resolved input and investigation of the amplitude for specific and relevant frequencies. Some of these analyses also act as event triggers for the sampling and storing of time resolved data at high time resolution before and sub-sequent the event. The same framework includes several solutions for the transfer of data and results to SCADA systems, historians and general severs. The data produced and gathered will serve as a basis for the evaluation of the current state and trends for wear and tear, but also for developing guidelines for more optimal power production.

Fig 2- Illustrating the hydropower plant-data value chain. FDB contributes with plant ground work, but also assists in visualizing and presenting data for decision makers at the power company.

Using time-resolved data as boundary conditions in numerical models of the water conduits, shaft-systems or for auxiliary systems such as fluid power systems, expand the knowledge of the hydropower unit or system performance. FDB deploys physical models on edge-processing units and in effect establish virtual instruments that can produce additional and useful information about the hydropower system. An alternative to edge processing involves the transfer of data from hydropower plants to servers running commercial and specialized physical modelling software. The result of these calculations when compared with independent plant data, or data not used as input, acts as a check for normal behavior and can ultimately detect an unwanted state. An established correspondence between this digital model and real plant data gives a confidence and such that the same model or the digital twin can be used for what-if scenarios and for troubleshooting. FDB uses expert user software and have the experience and capability of bringing these tools online.

FDB brings their expertise and experience related to digitalization in an industrial environment into the AFC4Hydro project. Combining this expertise with that of the other partners creates a formidable and very knowledgeable team that can manage most challenges. An important task in AFC4Hydro is to provide maximum knowledge of a unit or system state, but with minimum sensor- usage. This approach requires a superior physical understanding of the systems in question, access to and user knowledge of high level physical modeling software, knowledge of sensors and how to install and implement them in an industrial environment, and edge processing solutions that allow proper analysis of acquired data and control of physical systems to be validated through the AFC4Hydro project.


AFC4Hydro attended the 72nd American Physics Society (APS)- Division of Fluid Dynamics (DFD) meeting

Seattle Nov 24-26, 2019​

By Morten Kjeldsen, Managing Director Flow Design Bureau AS, Norway.

AFC4Hydro’s Morten Kjeldsen from FDB attended the 72nd APS-DFD meeting. A motivation for AFC4Hydro is to get inspiration and learn from flow control sessions at this meeting.  One of the impacts of AFC4Hydro project is to disseminate knowledge about flow control techniques and technologies both generally and for hydropower specifically. By attending this meeting the project also gets a first hand experience of a typical arena for presenting findings from AFC4Hydro.

The APS DFD meeting is an annual meeting that takes place in different cities in the US. More than 3000 persons attend these meetings. Most attendants are usually affiliated universities, where the majority originate from the USA but with a fair share from Europe. The duration corresponds to three full days, with as much as 40 parallel sessions. This meeting is also known for its Gallery of Fluid Motion that consists of both poster and video- entries. A reception is also held for the participants and this year the Museum of Flight hosted the event.

The general sessions are dominated by oral presentations of 12 minutes including a 2 minutes discussion. This format forces a presentation with a more focus on motivation, background, status of research and preliminary results. The consequence being that this meeting gives an insight into the status of the various disciplines of Fluid Dynamics. In addition to the general sessions the APS DFD includes invited talks, award lectures and flash-presentations of the abovementioned Gallery of Fluid Motion. The full meeting schedule is available through the conference web site.

During the meeting Kjeldsen went to the following sessions related to flow control: Separated Flows: Control, Drag Reduction, Flow Control: Sensor Placement and Shape Optimization, Aerodynamics: Flow Control, Flow Control: Separation and Flow Control: Turbulence/ Flow Control: Model Reduction. The only contribution that shared similarities with the technologies developed by AFC4Hydro was that by An et al: Dynamic mode analysis and control of vortical flows in pump sumps. The presentation showed an axial jet shooting from the floor of the pump sump and towards the pump intake. This solution mitigates the effects of the swirl experienced by the pump.

@Washington State Convention center in Seattle. Site of the 72nd APS DFD meeting.

Fig 1- @Washington State Convention center in Seattle. Site of the 72nd APS DFD meeting.

In addition to application of Flow Control, the technologies themselves where scrutinized. Passive solutions, such as shark-skins, were discussed. Semi active control using synthetic jets, suction and blowing were discussed but mostly in an aerodynamic context. These technologies are analogous to those exploited in the AFC4Hydro project. Of special interest was the use of Fluidic Oscillators, which is a way to produce oscillations in the angle of injection for the jets or sweeping jets. This approach is something the AFC4Hydro will evaluate and in relation to use with Hydropower.

The days spent at the meeting proved to be as expected. Lots of impressions and inputs that will benefit the work on AFC4Hydro. As for recommendations of the APS-DFD meeting; if you ever want to be overwhelmed by the share amount of work going on in the field of fluid dynamics, this is the place to go!


Link for technical program: http://meetings.aps.org/Meeting/DFD19/Content/3770

Fig 2- Reception at the Museum Flight. A European contribution, the Concorde, on display.