Development of Stand Alone Liquid Filled Miniature Power Generation System

A computational study is carried out to apprehend the flow characteristics of the standalone, liquid filled, miniature power system. A mesoscale combustion based thermoelectric power generator was designed as an alternate to the electrochemical batteries. This standalone system treat the walls of combustors as an energy source which in turn evaporate the liquid fuel stored in a surrounding tank. This vaporized fuel is then supplied to the combustor. The fuel is entrained through a high momentum fuel (vapor) jet, leading to the formation of combustible mixture with air. This partially mixed fuel/air mixture is fed to a mesoscale combustor and the flame is stabilized by facilitating hot gas recirculation regions.

A CFD simulation is carried out to understand the flow of the fuel form the tanks to the combustion chamber. A k-epsilon turbulent model clubbed with eddy dissipation is used along with Species transport mechanism. The fuel considered is ethyl alcohol. The fluid motion is taken to be inspired from Falkner – Skan flows.

INTRODUCTION

With the increase in production of small scaled power consuming devices, the demand of miniature power generation system has also increased. On one hand the electrochemical batteries are easily available but they have suffer disadvantages because of 2 reasons –

• • The low energy density

  • Longer recharging time

The liquid hydrocarbon fuel when compared to electrochemical batteries have a very high energy density. These fuels have a typical energy density of about 44 MJ/kg (12200 Wh/kg) while the most advanced rechargeable lithium/ion batteries have an energy density of just about 0.72 MJ/kg (200 Wh/kg) [1].

Note that for a fair comparison with electrochemical batteries, the electrical conversion efficiency and the mass of the energy conversion system also needs to be considered while estimating the energy density of a combustion based power generation system. (Reference of sir paper).

In the past , various miniature power generation systems have been constructed which uses gas turbine engines [3], rotary engines [4], free piston engines [5], reciprocating engines [1], and thermoelectric [7], thermos photovoltaic [8] based systems. Earlier examples include Hydrogen fueled gas turbine engine designed in MIT [3], miniature Wankel engine designed in UCB [4]. A butane fueled engine designed at University of Michigan [5]. We aim to use thermovoltaic or thermoelectric system as to reduce the amount of friction and leakage problem, because these systems will directly convert heat into electricity without the involvement of any moving parts. Although liquid fuel is much more preferred because of easy storage and refill capabilities but the field lacks in extensive amount of research.

Basically a standalone system is such that it does not have any moving parts and one of the easiest way to manufacture it is by facilitating fuel supply which when integrated with the system and air forms a combustible mixture. The mixture can be burned in a mesoscale combustor over which the hot side of thermoelectric modules can be mounted. The heat needs to be rejected from the cold side of the module and this needs to be done without using any external cooling system. The objective of the present study is to design a standalone, liquid fueled, mesoscale combustion based power generator with a power output of 15 W and endurance of 1 hour. The combustor is defined to be “mesoscale” if the physical length scale is in the range of 1 mm to 1 cm [1].

The prototype is designed to produce electrical energy using the expelled hot products as the heat source to a thermo-electric cell which works on the principle of Seebeck effect, where the temperature difference between the two junctions of the two electrodes produces an electro-motive force. The heat transfer from the walls of the combustion chamber is further used to evaporate the liquid fuel which then flows in as a jet and mixes with the entrained air. A cfd simulation of the combustion chamber was performed to augment the knowledge of the desired thermoelectric material that will be required on the top surface of the system. Since the combustion products directly hit the top surface, so principles of Falkner-Skan flows are used to model it. The details of which are covered further.

FIG2. AXISYMMETRIC MODEL FOR COMBUSTION CHAMBER AND DIMENSIONS OF FUEL INLET PIPE

Figure 3 displays the two variant zones of meshing, a fine mesh inside the pipe and a coarser mesh in surrounding region. The details of the meshing are given below –

No. of nodes- 497490

No. of elements – 494387

Max. skewness- 0.66

SOLUTION METHOD / MODELS

All the computations are performed using ANSYS (FLUENT) version 18.1. The computations are carried out with incompressible flow in a frame of reference attached to the model and the steady solution to the Reynolds Averaged Navier Stokes (RANS) equations is sought. A finite volume method capable of close to second order discretization, in primitive variables formulation, is employed. The SIMPLE algorithm is used for pressure-velocity coupling.

Standard k-epsilon viscous model is used for flow interpretation with default model constants.

Simple species transport mechanism with Volumetric reactions and ethyl-alcohol-air mixture material is used. The following equation governs the working of the

R_i is the net rate of production of species i by chemical reaction and S_i is the rate of creation by addition from the dispersed phase and the user defined source. J_i is the diffusion flux of species i, which arises due to concentration gradients and differs in both laminar and turbulent flows. The net source of chemical species i due to reaction, R_i which appeared as the source term in the species transport equation is computed as the sum of the reaction sources over the N_R reactions that the species participate.

In the analysis 2 turbulent chemistry models are used –

Finite rate model (for first 1000 iterations)

The laminar finite rate model computes the chemical source terms using the Arrhenius expressions and ignores turbulence fluctuations. This model provides with the exact solution for laminar flames but gives inaccurate solution for turbulent flames, in which turbulence highly affects the chemistry reaction rates, due to highly non-linear Arrhenius chemical kinetics. However this model may be accurate for combustion with small turbulence fluctuations example supersonic flames.

Eddy dissipation (for iterations ahead of 1000)

The eddy dissipation model, based on the work of Magnussen and Hjertager is a turbulent-chemistry reaction model. Most fuels are fast burning and the overall rate of reaction is controlled by turbulence mixing. In the non-premixed flames, turbulence slowly mixes the fuel and oxidizer into the reaction zones where they burn quickly. In premixed flames the turbulence slowly mixes cold reactants and hot products into the reaction zones where reaction occurs rapidly. In such cases the combustion is said to be mixing-limited, and the complex and often unknown chemical kinetics can be safely neglected. In this model, the chemical reaction is governed by large eddy mixing time scale. Combustion initiates whenever there is turbulence present in the flow. It does not need an ignition source to initiate the combustion. It is good for the non-premixed combustion but for the premixed flames the reactant will burn at the moment they enter in the computation model, which is incorrect. The reactant need some time to get to the ignition temperature to initiate the combustion. It is a shortcoming of this model.

Flow equation and heat transfer coefficient

Flows across a wedge type structure are characterize by a special name, termed as Falkner- Skan flows (fig4). They developed a similarity transformation method in which the partial differential boundary-layer equation was reduced to a nonlinear third-order ordinary differential equation which could then be solved numerically.

FIG 4 . A) FLOW CONFIGURATION OF FALKNER-SKAN FLOW, B) REPRESENTATION OF STAGNATION POINT

For a generalized wedge angle (βп) with proper assumptions we have

Where K is some constant, U is velocity far away and L is the overall length of the plate

Since in our case we want an a flow to hit the plate perpendicularly, the wedge angle = п

Therefore, βп = β.

Hence, β = 1 and m =1; which is termed as 2 D stagnation flow.

Therefore, it can be written as U_e(x) = cx …………. (iv)

Where c =

Using the heat transfer law for a plane plate, the local Nusselt number is obtained as :

This shows that The heat transfer coefficient is not a function of distance along the wall and assumes the same value throughout. This is the important factor that will play an important role in determining the type of photovoltaic material that will be used at the top.

BOUNDARY CONDITIONS

Following are the conditions used-

• Mass flow rate of fuel entering = 1.75 e-5 kg/m3

• Pressure inlet and pressure outlet conditions are used for air entry and exit.

• The top wall is patched with Aluminum Silicate and is open to convection with air

• The bottom portion is defined as axis that is needed because of axisymmetric modeling

• Remaining all other walls are kept adiabatic with Aluminum as the constituting material

• Mass fraction of fuel is unity for Ethanol

• Mass fraction for air is 0.23 for oxygen

MATERIAL PROPERTIES

AIR -

• Gage Pressure – 0

• Temperature – 300 K

• Density – 1.2kg/m³

ETHYL ALCOHOL-

• Density – System specified (incompressible-ideal-gas)

• C_p – System specified (mixing-law)

ALUMINUMN SILICATE –

• Density – 3203 kg/m3

• C_p – 164.87 + 36.97xT – 4.28xT² +0.69xT³

TEMPERATURE PLOTS

The bottom most portion has lower temperature because of the air entrainment which is at lower temperature of 300 K. The flame gets stabilized in the narrow vertical passage. The flame temperature corresponding to this case is near about 2360 K. This temperature is in agreement with the adiabatic flame temperature of 2355 K.

Stoichiometric equation –

C₂H_5­OH + 3(O₂ + 3.76N₂) + 2CO₂ + 3H₂O + 11.28N₂

Air fuel ratio – 8 approximately (theoretical)

Cfd Air fuel ratio used – 4.5

So equivalence ratio – 4.5/8 = 0.56, this imply that the mixture is lean

It can be easily seen from the figure 6 that first three 25% regions of the top surface (from the centerline) have a decrease in temperature of about 200K each whereas there is a rapid decline of temperature in the last ¼ the region. This is mainly because of the presence of the outside air which cools down the surface upon contact with it.

VELOCITY PLOTS

The combustion chamber is so designed in a manner that it will ensure choked flow at the port, with downstream pressure equal to critical pressure. Note that there might be some pressure loss in the combustion chamber but it is expected to be very less and can be avoided. At the choked condition the temperature can be calculated as

T/T0= 2/(1+γ)

where gamma = 1.13 and To = 365.8

This will give, T = 342.99 K which is the inlet temperature of the fuel entering the chamber

The exit velocity corresponding to it can be calculated as

Vexit = sqrt(2rTo/(1+γ))

this comes out to be approximately 264.9 which is in close approximation to the simulated velocity.

The jet decays after approx. at a distance corresponding to 12 times the diameter of the exit

Fig 9 clearly shows the development of fuel jet in the combustion chamber. The axial velocity keeps on decreasing until it stabilizes. Apart from the presence of the outlet chamber, recirculation of flow is another main reason because of which the temperature in the top chamber decreases. These recirculation zones help is distribution of heat energy thereby reducing the local temperature.

CONCLUSIONS

Combustion accounts to 60% heat release in the chamber. Even after losing 60% heat of the heat, the temperature of the exit gases is expected to drop from 2300K to about 950K. But this temperature is still greater than the hot side temperature of the thermoelectric modules which usually work at around 773K. The aluminum silicate slip that we attached on the hot side of TEM will ensure uniform temperature distribution on the surface of TEM. Based on the temperature profile at the top, a variable thickness sheet of aluminum silicate can be designed.

REFERENCES

1. Ju, Y., Cadou, C., and Maruta, K., 2015. Microscale Combustion and Power Generation. Momentum Press, New York.

2. Ju, Y. and Maruta, K., 2011. “Microscale Combustion: Technology Development and Fundamental Research”. Progress in Energy and Combustion Science, 37, pp 669-715.

3. Epstein, A. H., Senturia, S. D., Al-Midani, I., Ananthasuresh, G., Ayaon, A., Breuer, K., Chen, K. Ehrich, F. F., Esteve, E., Freshette, L, Gauba, G., Ghodssi, R., Groshenry, C., Jacobson, S. A.,

4. Kerrebrock, J. L., Lang, J. H., Lin, C., London, A., Lopata, J., mehra, A., Mur-Miranda, J. O., Nagle, S., Orr, D. J., Piekos, E., Schmidt, M. A., Shirley, G., Spearing, S. M., Tan, C. S., Tzeng, Y. and Waitz, I. A., 1997. “Micro-Heat Engines, Gas Turbines, and Rocket Engines – the MIT Microengine Project”. AIAA Fluid Dynamics Conference, Snowmass Village, CO, June 29 – July 2.

5. Fu, K., Knobloch, A. J., Martinez, F. C., Walther, D. C., Fernandez-Pello, C., Miyaska, K. and Maruta, K., 2001. “Design and Experimental results of Small- Scale Rotary Engines”. ASME International Mechanical Engineering Congress and Exposition, New York, NY, Nov 11-16.

6. Dahm, W. J. A., Ni, J., Mijit, K., Mayor, R., Qiao, G., Benjamin, A., Gu, Y., Lei, Y. and Papke, M., 2002. “Micro Internal Combustion Swing Engine (MICSE) for Portable Power Generation Systems”. AIAA Aerospace Sciences Meeting, Reno, NV, Jan 14-17.

7. Huth, J., and Collins, J., 2007. “Diesel Fuel-to-Electric Energy Conversion using Compact Portable, Stirling Engine-Based Systems”, JSME International Stirling Engine Conference, Tokyo, Japan, Sep 23-26.

8. Yadav, S., Yamasani, P. and Kumar, S., 2015. “Experimental Studies on a Micro Power Generator using Thermo-electric Modules Mounted on a Micro- Combustor”. Energy Conversion and Management, 99, pp 1-7