A Spin-Off Company of A Comprehensive Method for the Characterization of Engine Heat Rejection Giuseppe Cicalese Stefano Fontanesi Fabio Berni
Genesis of the methodology Methodology guidelines In-Cylinder / CHT framework interaction CHT model setup Piston analysis Gas-to-wall thermal boundary conditions Thermal laws of the wall Approaches in literature A modified formulation Results Summary and conclusions Q&A Presentation Outline
Trends in SI engines Downsizing High specific power Limits due to Thermo-mechanical failures Genesis of the methodology Abnormal combustion onset Turbocharging Needs Engine thermal field evaluation Abnormal combustion prediction Detailed CFD analyses are useful for design improvements
Accurate thermal field representations and reliable estimations of the coolant heat rejection depend on Genesis of the methodology A detailed Conjugate Heat Transfer (CHT) model including all the solid components and the coolant circuit A proper set of boundary conditions, especially on the components facing the combustion chamber
es-ice TM & STAR-CD Methodology guidelines The in-cylinder and the CHT frameworks are strictly related: they are coupled in order to properly calculate the global amount and the spatial distribution of the thermal loads acting on the components facing the combustion chamber STAR-CCM+ Preliminar in-cylinder simulation with uniform wall temperatures Heat flux export and mapping on the CHT model New in-cylinder simulation with spatial-distributed wall temperatures Evaluation of the engine thermal field and extraction of the point-wise temperatures of the components facing the combustion chamber
Fluid region The computational domain covers the entire cooling circuit surrounded by block and head Core mesh constituted by polyhedral cells Prismatic cells next to the walls Particular care is devoted to the mesh of thin passages Methodology guidelines Solid regions Heads, block, liners, valves, guides, seats, gaskets, spark plug are included Piston is separately analyzed Core mesh is constituted by only polyhedral cells Prismatic cells are placed next to the walls only if the temperature dependency of the thermal conductivity is known A proper contact resistance is set between all the adjacent elements
A time-averaged point-wise heat flux distribution deriving from a 3D-CFD combustion simulation is applied on the piston crown The effects of oil jets, piston rings, frictions, lubricant and coolant circuit are simultaneously taken into account The resulting Q liner is then to the liner Ring Groove Methodology guidelines Piston Skirt
In a CHT model, all these effects have to be taken into account: Combustion Charge exchange Frictions Lubricating effects Environmental conditions Adiabatic effects Press fit contacts Pairs of heat transfer coefficient and reference temperature properly set Methodology guidelines Contact resistances between adjacent elements Combustion Friction of the piston ring Friction of the piston skirt Heat transfer from the piston
What is the right amount of heat transfer due to combustion? Gas-to-wall thermal BCs If experimental data deriving from engine thermal surveys are available, target heat transfer is known and the heat fluxes might be properly scaled in order to match the correct integral amount preserving the point-wise thermal distribution. Evaluated from engine thermal survey Otherwise, if target heat flux is not a-priori known, one has to rely on the values estimated via thermal laws of the wall but
Standard q w = ρc pu τ (T T w ) Pr t (u + + P) q w = q w = everyone says his own! Modified Angelberger ρc p u τ Tln T T q w = W Pr t (u + + P) Angelberger ρc p u τ Tln T T W 2.075lny + + 3.9 ρc p u τ Tln T W T dp dt + Q c 1 0.4767 ln y+ + Launder & Spalding Gas-to-wall thermal BCs ρc p u τ (T T w ) q w = 2.075ln(9y + ) + 7.52 Pr 0.85 1 0.85 Pr Rakopoulos ν y + 40 u τ 0.4767 + 1 + 117.31 Pr 1 Pr0.4767 ln 40 + 1 Pr0.4767 + 10.2384 Han & Reitz ρc p u τ Tln T T q w = W 2.1lny + + 2.5 0.25
Thermal laws of the wall Approaches in literature Most of the wall functions are validated using the GM pancake test engine, whose characteristics are far away from those typical of current production engines. Angelberger and Han and Reitz formulations are analyzed because of their large diffusion. Bore 105 [mm] Stroke 95.25 [mm] Connecting Rod Length 158 [mm] Compression Ratio 8.56 Engine Speed 1500 [rpm] Equivalence Ratio 0.87 Volumetric Efficiency 40 % Spark Timing 27 CA bftdc
Pressure [bar] Thermal laws of the wall Approaches in literature Heat Flux [W/m^2] Local wall heat flux measurements are experimentally available thanks to four probes located on the cylinder head. For the sake of brevity just one of them is shown. 25 20 15 In-cylinder Pressure Pressure 3D Pressure Exp 2 1.8 1.6 1.4 1.2 1 Heat Flux - HT1 Angelberger Han & Reitz Heat Flux Exp Both the Angelberger and the Han and Reitz models closely resemble the measured heat flux 10 0.8 5 0-30 -20-10 0 10 20 30 CA 0.6 0.4 0.2 0-30 -25-20 -15-10 -5 0 5 10 15 20 25 30 CA These laws of the wall are capable to well predict the heat fluxes in these operating conditions
Thermal laws of the wall Approaches in literature Angelberger and Han and Reitz laws of the wall are then applied to the in-cylinder simulation of three different engines in order to evaluate the heat transfer to the walls Engine A B/S 1.05 Operating Condition 1 Revving Speed 7000 rpm BMEP 18.33 bar Operating Condition 2 Revving Speed 5000 rpm BMEP 24.36 bar Engine B B/S 0.93 Operating Condition Revving Speed 5200 rpm BMEP 22.96 bar Engine C B/S 1.04 Operating Condition Revving Speed 7000 rpm BMEP 20.03 bar The calculated heat transfer is compared with the estimated one derived from the engine thermal survey, available for all the analyzed engines The thermal field is compared with experimental thermocouple measurements
Thermal laws of the wall Approaches in literature Target from Thermal Balance Average to CHT - Angelberger Average to CHT - Han and Reitz Engine A - Op.Cond. 1 1 1.8 1.6 1.4 0.6 1.2 1 0.4 0.8 0.2 0.6 0.4 0 0.2-0.2 0 90 180 270 Head Int. Port Target from Thermal Balance 360 450 CA 540 Liner Exh. Port Average to CHT - Angelberger 630 720 2 Cycle - Averaged Heat Transfer Boundary Heat Transfer 0.8 810 Piston Total Average to CHT - Han and Reitz 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 Engine A Op. Cond. 1 Engine A Op. Cond. 2 Engine B Engine C Angelberger wall function tends to highly overestimate the wall heat flux (+30% +50%) This behavior is confirmed for all the analyzed cases Han and Reitz wall function shows a higher overestimation A unique trend and/or scale factor can not be identified
Angelberger and Han and Reitz laws of the wall are somehow similar both in formulation and in principle: they take into account the variation of density and viscosity within the boundary layer and are actually a compressible version of the traditional engine wall treatments, in which gas θ+ 20 compressibility is not accounted for. Angelberger thermal law of the wall: θ + = Pr η + ; η + 13.2 θ + = 2.075 ln η + + 3.9; η + > 13.2 Han & Reitz thermal law of the wall: θ + = 2.1 ln η + + 2.5 Thermal laws of the wall Approaches in literature For both temperature profiles the wall heat flux formulation is given as: 18 16 14 12 10 ρ w c p u τ T w ln q w = where η + = (ν w /ν) y + is the non-isothermal dimensionless distance. θ + 8 6 4 2 0 Angelberger Han and Reitz 1 10 100 1000 η+ T T w
A similar law of the wall was already proposed by Kays and Crawford for isothermal (or incompressible) flow: and the wall heat flux was defined as: Thermal laws of the wall A modified formulation T + = 2.075 ln y + + 13.2 Pr 5.34 (T + = 2.075 ln y + + 3.9 for air @ Pr = 0.7) q w = ρ w c p u τ (T T w ) T + The alternative approach proposed here is somehow different: isothermal laws of the wall are applied to non-isothermal problems once again, but new scales of velocity, temperature and wall distance are here introduced for the inner zone of the boundary layer. Compared to Kays and Crawford s heat transfer model, where such scales are computed as now the new scales are u τ = τ w ρ w T τ = q w ρ w c p u τ u τ T τ y τ y τ = ν w u τ
Thermal laws of the wall A modified formulation These scales are used to obtain non-dimensional wall distance and temperature respectively: y + = y y τ T + = T T w T τ The resulting wall heat flux formulation is: q w = H (T T w ) T + The so-called isothermicity parameter ζ can be introduced as a characteristic scale of the ratio between the gas temperature and the wall temperature and as a characteristic scale of the heat flux (it is, in the end, a non-dimensional flux) in order to understand if a law of the wall is more suitable than another one ζ = T τ T W = q w ρ w c p u τ T w ζ 0 Heat Flux~0 ; T/T w ~1 Isothermal Approach ζ 0 Heat Flux 0 ; T/T w 1 Non-Isothermal Approach
ζ ζ ζ Thermal laws of the wall A modified formulation In order to understand why such models correctly match the test case measurements and overestimate the heat fluxes in the investigated engines, the evolution of the ζ parameter during the cycle is analyzed 0.4 0.3 0.2 0.1 GM Pancake Engine A - Op. point 2 Engine A - Op. point 1 Engine B Engine C ζ - Head 0.3 0.2 0.1 GM Pancake Engine A - Op. point 2 Engine A - Op. point 1 Engine B Engine C ζ- Liner 0.3 0.2 0.1 GM Pancake Engine A - Op. point 2 Engine A - Op. point 1 Engine B Engine C ζ - Piston 0-30 -20-10 0 10 20 30 40 50 60 70 80 CA 0-30 -20-10 0 10 20 30 40 50 60 70 80 CA 0-30 -20-10 0 10 20 30 40 50 60 70 80 CA Under the assumption of ideal gas, ζ can be written as follows: ζ = q w p u τ γ/(γ 1) being p the in-cylinder pressure and γ the ratio of specific heats. In production engines ζ is lower because of the much higher incylinder pressure and turbulence levels. In comparison with the pancake engine, the investigated production engines exchange less energy through the walls. Heat transfer models by Angelberger or Hand and Reitz may correctly predict gas-to-wall heat fluxes for high ζ values, as in the pancake case, while they tend to overestimate fluxes for low ζ values
Results 1/2 2 Thanks to the new formulation of the thermal wall function, the CHT model is able to well match both the global engine thermal survey and the local temperature field 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 Engine A - Op. Cond. 1 Engine A - Op. Cond. 2 Target from Thermal Balance Average to CHT - Han and Reitz Engine B Engine C Average to CHT - Angelberger Average to CHT - R&D CFD
C C C C Thanks to the new formulation of the thermal wall function, the CHT model is able to well match both the global engine thermal survey and the local temperature field 260 240 220 200 180 160 140 120 100 260 240 220 200 180 160 140 120 100 Engine A Op. 1 0 5 10 15 20 # Thermocouple Experimental R&D CFD Engine B - Head 0 10 20 30 40 # Thermocouple Experimental CFD Modified 260 240 220 200 180 160 140 120 100 220 200 180 160 140 120 Results 2/2 Engine A Op. 2 0 5 10 15 20 # Thermocouple Experimental R&D CFD Engine B - Block 100 40 50 60 70 80 # Thermocouple Experimental R&D CFD
Summary & conclusion A brief overview of the integrated in-cylinder / CHT approach for heat transfer modeling in internal combustion engine has been given in order to take into account all the relevant phenomena which could occur in an internal combustion engine Angelberger and Han and Reitz laws of the wall formulations have been applied to in-cylinder simulations of the GM pancake test engine, and the agreement with the experimental data has been proved to be very good The same thermal wall treatments has been applied to three DISI turbocharged high performance engines in order to evaluate the thermal loads acting on the components facing the combustion chamber, with a resulting overestimation of the heat fluxes A modified law of the wall formulation has been proposed and tested on the same engines and operations, and a good agreement has been found both in engine thermal survey and in temperature distribution
A Spin-Off Company of Thank you for your attention! Any questions? Giuseppe Cicalese, PhD giuseppe.cicalese@red-cfd.it Prof. Stefano Fontanesi stefano.fontanesi@unimore.it Fabio Berni, PhD Student fabio.berni@unimore.it