Shock/turbulence interactions in dense gases

Lead Research Organisation: Imperial College London
Department Name: Mechanical Engineering

Abstract

To reduce the UK's greenhouse-gas emissions anywhere near the legally-binding 2050 targets, a major attack on both energy wastes and unsustainable forms of electricity production is essential. Owing to their appealing thermo-physical properties (e.g. large heat capacity relatively to the molecular weight, low boiling point, elevated density), molecularly-complex and dense gases (e.g. hydrocarbons, perfluorocarbons, siloxanes) are at the heart of realistic solutions for thermal power stations to operate efficiently on low-temperature heat sources (e.g. solar, biomass, geothermal), where they are used as substitute for water steam (e.g. organic Rankine cycle). Flow expanders in such power stations partially operate in the vicinity of the thermodynamic critical point, where the speed of sound is substantially reduced, turning the expander flow into a highly supersonic gas flow, inevitably leading to the formation of shock waves.

Shock waves have the detrimental property of degrading the expander efficiency by dissipating kinetic energy into heat, and by promoting viscous losses through boundary-layer separation and thickening. Quite remarkably, and contrary to ideal gases, shock waves in molecularly-complex and dense gases can be made almost isothermal, therefore relieving part of the efficiency losses imparted by the shock wave. This remarkable property is a direct consequence of the exceptionally large number of active degrees of freedom of the gas molecule. While the prospect of efficient supersonic expanders is appealing, little is known on the implication near-isentropic shocks have on the amplification of turbulence fluctuations (which are always present in turbines). In particular, shock/turbulence interactions in dense gases can lead to the emission of energetic acoustic waves, which are significantly more powerful than in standard ideal gases. If present, such acoustic forcing can erode the expected turbine efficiency, generate vibrations and cause premature blade fatigue. The proposed research will establish a robust and fundamental understanding of sound emission from shock/turbulence interactions in dense gases, and provide a new understanding of the underlying physics, which will allow the development of predictive tools that can inform future design choices.

Planned Impact

The Earth's surface receives enough solar radiation in 10 minutes to fulfill the world's annual electricity consumption. Yet, the actual contribution of solar energy to the mix is below 1%. One particular technology, known as thermal solar power (TSP), uses the sun's radiation to heat up the working fluid of a steam turbine, a method that could produce 20% of Europe's electricity consumption by 2050, including in the UK. Using molecularly-complex and dense gases as working fluids in the turbine would enable TSP power stations to run on low-temperature heat source, a technology which could dominate the kW to MW power-output market. However, emission of strong acoustic waves from the shock/turbulence interactions, inevitably leading to varying inlet-flow angles and vibrations of the turbine, can cause premature blade fatigue and halve the turbine efficiency and the solar radiation to electricity conversion currently assumed. Developing a tool to avoid triggering such vibrations is therefore invaluable to TSP manufacturers, but also to energy policy makers so as to ensure that their assumptions on expected energy performances are as accurate as possible. Therefore, in addition to the aforementioned academic beneficiaries, software developers in the area of computational fluid dynamics (CFD) are directly targeted by the proposed research, and this is reflected by the partnership with a leading CFD company. Developing a low-cost and reliable technique to capture and predict possible acoustic emission from the shock, easily fitting in standard CFD packages, would be extremely valuable to turbine designers. Ultimately, power station operators, their customers, and environmental policy makers, would all benefit from the availability of more efficient and robust turbine designs.

Publications

10 25 50

publication icon
Alferez N (2017) Shock-induced energy transfers in dense gases in Journal of Physics: Conference Series

publication icon
Touber E (2019) Shock-induced energy conversion of entropy in non-ideal fluids in Journal of Fluid Mechanics

publication icon
Touber E (2019) Small-scale two-dimensional turbulence shaped by bulk viscosity in Journal of Fluid Mechanics

 
Description Dense gases are characterised by molecules featuring large numbers of active degrees of freedom (quantified by a large heat capacity relative to the molecular weight). Isentropes in such gases have the very distinct property of following rather closely the isotherms. The so-called critical isotherm is well known to feature a plateau at the liquid-vapour critical point (in the pressure/specific-volume phase diagram). Dense-gas isentropes will therefore inherit such property and be able to "bend" towards a plateau near the liquid-vapour critical point. From an application point of view, this is very interesting: indeed, shallow isentropes offer the possibility of achieving large expansions for virtually no pressure drop - this is the ultimate goal of a turbine (which is designed to expand a gas as isentropically as possible). However, with shallow isentropes come reduced sound speeds and extreme compressibility. This project has investigated whether such an extreme compressibility can modify how velocity fluctuations (inherently present in turbomachinery, e.g. turbulence) evolve and more specifically behave when interacting with shock waves (also inevitably present in such expanders). Below are some of the most remarkable results.

o What were the most significant achievements from the award?
[Theory] It was found that the shock adiabat, which represents all possible post-shock states given a pre-shock state, also closely follows the isotherm. In the vicinity of the critical point, this leads to two major results:
- First, the relatively shallow adiabat (in pressue/specific-volume phase) can increase the transmission coefficients of an incoming perturbation associated with density fluctuations (entropic and acoustic) by more than two orders of magnitude relative to an ideal gas.
- Second, the presence of a near-plateau on the adiabat makes the post-shock state multi-valued. However, only one solution is found to be observed (referred to as being admissible). This makes the adiabat discontinuous. Whilst this is not a new result by itself, we show that the transmission and generation coefficients at the shock have singularities associated with the discontinuous adiabat. This is a new theoretical result which is significant as it clearly establishes that arbitrarily similar flow conditions can lead to major differences in the post-shock fluctuations. In realistic flows, slight variations of the operating conditions are inevitable and could therefore trigger very different post-shock fluctuations.

The above results have been derived analytically assuming the incoming perturbation to be very small compared to the time-mean flow properties - this is the linear theory. However, a flow solver was also developed to be able to comment on non-linear interactions. In particular, the simulations have confirmed the possibility of re-directing the energy of an incoming perturbation into strong density, acoustic or shear waves almost independently of each others. This is simply impossible to achieve in an ideal gas. This result is of interest as it suggests that the pre-shock turbulence kinetic energy can indeed by redirected into strong acoustic waves behind the shock, but that such conversion may be very selective on the basis of the incoming Mach number.

[Tools] A new high-fidelity solver to simulate compressible flows featuring arbitrary equations of states has been developed. The solver will be applied in the context of compressible aerodynamics for aerospace and energy applications and represents a major step forward. This is significant since current high-fidelity solvers assume the gas to be ideal and this assumption is used extensively when implementing the numerical method, making it difficult to extend their use for non-ideal gas. The new in-house solver developed during this project does not rely on such assumptions and can address important questions relating to real-gas aerodynamics and the new avenues by which turbulence evolve. In particular, a new shock-capturing method was developed which relies on analytical solutions to the viscous shock structures and offers a good control on the resolvability of shocks. This new approach could be also beneficial to codes based on the ideal-gas assumption.

o To what extent were the award objectives met? 
Two main objectives were sought.
First, to what extent do dense-gas effects modify the transfer of turbulence kinetic energy into acoustic waves at the shock? The work undertaken has clearly shown that dense-gas effects can significantly modify the refracted energy. Some shocks were found to be able to generate acoustic waves two orders of magnitude more powerful than that of an equivalent ideal gas. Some shocks were also found to be able to completely damp acoustic waves. Moreover, such differences in the refraction properties (from strong to no acoustic waves) were found to be able to occur over a very small modification in the pre-shock conditions, thus highlighting the extreme sensitivity of the shock-refraction properties. This was explained analytically (and related to the discontinuous nature of the admissible shock adiabat) and confirmed numerically using a newly-developed flow solver (both for a very small (linear) and regular (non-linear) incoming perturbation).
Second, can inertial effects arising from inter-molecular forces at the shock promote new hydrodynamic instabilities, and change the post-shock turbulence intensity? The stability analysis carried during the project did not include the viscous structure of the shock, but the numerical simulations did. We have been able to confirm numerically that the viscous-shock structure does not modify the theoretical predictions based on inviscid (zero shock thickness) considerations, provided that the wavelength of the perturbation is much larger than the actual shock thickness. In particular, we have provided strong evidence for the existence of the so-called D'yakov-Kontorovich instability, even in the presence of viscous-shock structures. This is also a new result and more work is needed to be able to comment on the impact this may have on the post-shock turbulence intensity though.
Exploitation Route The above findings have highlighted some of the unique properties associated with dense gases undergoing extreme compressibility effects. None of these properties are currently accounted for in turbulence models used by industry. The immediate step forward from the work is to study turbulence in such flows and work closely with software companies providing simulation tools (CFD) such as the partner on this project (i.e. CD-adapco).

The ability to re-direct large amounts of turbulence kinetic energy into acoustic waves and internal energy for very slight variations in operating conditions should be of interest to turbine and compressor designers exploiting the unique thermo-physical properties of the liquid-vapour critical point, not only to avoid unwanted but potentially energetic perturbations but perhaps as a means to achieve a new kind of flow control.
Sectors Aerospace

Defence and Marine

Energy

Transport

 
Description Invited Professor at ENSAM ParisTech
Amount € 2,000 (EUR)
Organisation ParisTech Arts et Metiers 
Sector Academic/University
Country France
Start 05/2016 
End 07/2016
 
Description Invited Professor at OIST
Amount £5,000 (GBP)
Organisation Okinawa Institute of Science and Technology 
Sector Academic/University
Country Japan
Start 12/2016 
End 01/2017
 
Description Short-term invitation fellowships
Amount ¥700,000 (JPY)
Funding ID S15079 
Organisation Japan Society for the Promotion of Science (JSPS) 
Sector Public
Country Japan
Start 03/2015 
End 04/2015
 
Description OIST 
Organisation Okinawa Institute of Science and Technology
Country Japan 
Sector Academic/University 
PI Contribution Dr Chakraborty and his team investigate fundamental aspects of turbulence using gravity-driven soap films. In the process, they have discovered the existence of Marangoni shocks in such experiments (Tran et al. PRL, 2009). Such discovery is important as Marangoni shocks are easily (and inexpensively) studied compared to their gas-dynamic counterparts. This allows us to study how turbulence is refracted across a shock, which is the ultimate aim of the EPSRC first grant, albeit in dense gases. The collaboration with OIST consists in modifying the high-fidelity code developed under the EPSRC grant to simulate the Marangoni shock. We therefore bring our expertise on numerical simulations of shock/turbulence interactions. The overall objective is to explain the observed increased turbulence intensity behind the Marangoni shock.
Collaborator Contribution This collaboration gives us access to an experimental setup to study a similar process that the one we study numerically and theoretically. This is a great opportunity to apply and validate our theory on an actual configuration. OIST also gave us access to their supercomputer, allowing us to run our simulation software on thousands of cores.
Impact So far, the outcome has been to secure funds through JSPS to visit OIST. The PDRA employed on the EPSRC 1st grant was also given the opportunity to travel twice to OIST for a total of 4 months. Scientific output is the immediate aim and we hope to be able to report those in the near future.
Start Year 2015
 
Description Heat Recovery Workshop 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Industry/Business
Results and Impact Experts from the energy sector attended. Our contribution was to present some of the peculiarities of real-gas aerodynamics in turbomachinery. The talk was well-received and allowed us to be asked to lead a group on "fundamentals of non-ideal compressible fluid dynamics" at the first conference on this topic (NICFD 2016).
Year(s) Of Engagement Activity 2015
URL http://heat-recovery.eng.cam.ac.uk/WebHome
 
Description ICTAM 2016 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact Every four years, the mechanics community from around the world gathers at the Congress of Theoretical and Applied Mechanic organised under auspices of the International Union of Theoretical and Applied Mechanics (IUTAM) to discuss science and strengthen relationships in the broad area of mechanical engineering. We have presented our findings on the unusual refraction properties of shocks in non-ideal substances. The presentation was well received and has led to discussions, invitations in research laboratories to discuss possible experimental validation of the results, as well as requests from journal editors to submit our findings in their journals. The broad nature of the meeting also meant that we could identified other applications areas not included in the initial proposal (biology and transportation).
Year(s) Of Engagement Activity 2016
URL http://www.ictam2016.org/welcome_e.shtml
 
Description ISROMAC16 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact The International Symposium on Transport Phenomena And Dynamics of Rotating Machinery is a conference covering areas like Aero-acoustics, Rotor dynamics, Hydraulic Machinery, Automotive Turbomachines, Energy converters, Boundary layer flows, Turbulence modelling, CFD in Turbomachinery, Combustion, Jet engines and much more. Our participation was important in order to make the turbo-machinery community aware of possible and unexpected flow phenomena in components operating near the liquid-vapour critical point associated with the unusual shock-refraction properties. Some aspects of the phenomena have been discussed in private in a subsequent meeting with a world-leading manufacturer of waste-heat-recovery systems.
Year(s) Of Engagement Activity 2016
 
Description NICFD2016 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact The 1st International Seminar on Non-Ideal Compressible-Fluid Dynamics for Propulsion & Power (NICFD 2016) was a gathering "intended to promote the exchange of scientific information, to encourage and consolidate the interaction between researchers and professionals, with a special emphasis on the progress in research, development, and applications of the topics related to the field of propulsion and power". We (once again) have reported some unusual aspects of shock-refraction properties in non-ideal gases. However, this time we have shown the audience results using state-of-the-art equations of state instead of the toy ones we had used previously. By demonstrating that significant real-gas effects were indeed expected in substances used inside actual devices such as turbines (not just in toy problems), we have convinced the audience about the relevance of the phenomenon we had discussed in the past year.
Year(s) Of Engagement Activity 2016
URL http://easychair.org/smart-program/NICFD2016/index.html