Professional Documents
Culture Documents
Simulation of Wankel 2012
Simulation of Wankel 2012
It may not be reproduced, stored in a retrieval system, distributed or transmitted, in whole or in part, in any form or by any means.
Downloaded from SAE International by Leonid Tartakovsky, Tuesday, October 16, 2012 03:48:23 AM
Jonathan Reif
Elbit Systems Ltd
cylinder, 4-stroke RP engine. The paper does not present MAIN DIFFERENCES BETWEEN
criteria for choice of the virtual RP engine dimensions, heat
transfer and heat release calculations. In the work [16] the WANKEL AND PISTON ENGINES
commercial software GT SUITE was used for prediction of The following main differences between Wankel and piston
the Wankel combustion chamber pressure as a function of engines that affect their performance can be mentioned:
rotor angle. As in [15], the simulated Wankel engine was
replaced with the virtual 3-cylinder, 4-stroke RP engine. • Difference in patterns of working chamber volume and
Some assumptions that do not reflect peculiarities of Wankel surface dependence on the angle of shaft rotation;
engine geometry and working cycle were made in [15]: heat • Duration of the Wankel working cycle is 1.5 times longer
release and heat transfer calculation procedures were used compared with a 4-stroke piston engine (in terms of the angle
unmodified and did not take into account changes of of shaft rotation). Complete working cycle of a Wankel
combustion and heat transfer processes in Wankel engine engine takes place in each working chamber per one rotor
compared with RP counterpart, piston stroke of the virtual RP revolution, or per three shaft revolutions. In other words, one
engine was taken equal to eccentric offset of the simulated working cycle in a Wankel engine occurs during 1080
Wankel engine, etc. These simplifications do not allow degrees of shaft rotation (or 360 degrees of rotor rotation),
prediction of Wankel engine performance with an acceptable compared to 720 degrees of the crankshaft rotation in a 4-
accuracy. stroke piston engine;
The approach of [15 and 16] with use of the GT-POWER • “Hot” and “cold” stator zones of the Wankel working
software of the GT-SUITE package is discussed and further chamber surfaces are separated contrary to a piston engine,
developed in this paper. The GT-SUITE is an up-to-date where the same working chamber surfaces are heated and
integrated program for engines simulation and design cooled in-turn. This, in combination with charge rotational
analysis. It contains the following components: GT-POWER movement together with a working chamber, leads to
- engine simulation for performance and acoustic analysis; differences in the heat transfer conditions;
GT-DRIVE - vehicle performance and cycle analysis for fuel
economy and emissions; GT-VTRAIN - valve train • Unfavorable shape of the Wankel working chamber (Figure
kinematics, dynamics and tribology; GT-FUEL - injection 1) leads to the higher surface-to-volume ratios and larger
system pressure and flow analysis; GT-COOL - engine heat relative value of crevice volumes, where flame quenching
and cooling system analysis; GT-CRANK - crankshaft takes place. This causes differences in combustion patterns;
dynamics and torsional vibrations analysis. All six • Wankel working chamber has more complicated design of
components of GT-SUITE are based on a common set of seals (apexes and rotor side seals). This leads to the
multi-physics libraries and thus have a large degree of possibility of higher charge leakage values;
commonality among them. The GT-POWER is the 2- and 4-
stroke SI and diesel engine simulation tool widely used • Wankel engine differs fundamentally from piston engine in
nowadays by engines makers. Among its advantages are ease its kinematic mechanism. This leads to differences in internal
of use and tight integration with the rest GT-SUITE friction power losses.
components. Its flexibility and usefulness are further
enhanced by the possibility of integration with STAR-CD,
Fluent, Simulink, MS/EXCEL and other general purpose
software. Due to the large size of its user base, the GT-
POWER has been thoroughly validated by real-life
applications. Use of this software which is focused on RP
engines for simulation of Wankel engine performance has
required development of the algorithm for composing a 3-
cylinder, 4-stroke virtual RP engine that would be
geometrically similar to the considered Wankel engine.
Development of special procedures that would take into
account differences between Wankel and RP engines in
combustion and heat transfer, as well as assessment of
discharge coefficients of intake and exhaust ports, was Figure1. Unfavorable shape of the working chamber in a
required too. Wankel engine [17].
THIS DOCUMENT IS PROTECTED BY U.S. AND INTERNATIONAL COPYRIGHT.
It may not be reproduced, stored in a retrieval system, distributed or transmitted, in whole or in part, in any form or by any means.
Downloaded from SAE International by Leonid Tartakovsky, Tuesday, October 16, 2012 03:48:23 AM
SIMILARITY CRITERIA volume ratio for the VP engine should not differ from the
corresponding values of the modeled Wankel engine by more
The method of Wankel engine performance simulation by then 1% at minimal working chamber volume values of these
using RP engine software which is discussed in this paper engines.
requires compilation of some similarity criteria for these
engines and composing a virtual piston (VP) 4-stroke engine A flow chart of the developed algorithm for calculation of the
that would allow simulation of Wankel performance. geometric similarity parameters is presented in the Appendix
B. Results of the computations are shown in Tables 1, 2 and
The following similarity criteria of Wankel and the VP Figures 4 - 6 for the example of the 802W Wankel engine. Its
engines have been defined: main parameters are:
• Displacement equality;
• Naturally aspirated, spark ignition, single-rotor, combined
• Compression ratio equality; air-liquid cooling;
• Identical behavior of the working chamber volume • Rated shaft speed - 8,000 rpm;
dependence on angle of shaft rotation;
• Rated brake power - 52 kW;
• Identical behavior of working chamber surface-to-volume
ratio change vs. angle of shaft rotation; • Displacement of each working chamber - 343.9 cc;
• Identical behavior of intake and exhaust ports discharge • Compression ratio - 7.4;
coefficients vs. angle of shaft rotation.
• Eccentricity - 11.6mm;
In addition to these similarity criteria which may be • Generating radius - 71.5mm;
summarized as a geometric similarity, the following
important aspects should be taken into account: • Rotor width - 80mm.
• Instantaneous values of the heat transfer through the As can be seen from Table 2, a maximal deviation from the
combustion chamber walls vs. angle of shaft rotation have to geometric similarity at the top dead center (TDC) does not
be calculated separately for non-firing (“cold”) and firing exceed 0.8% and at the bottom dead center (BDC) - 3.5%. It
(“hot”) parts of the working cycle; is clear that from the viewpoint of combustion and
• Modeling the combustion should take into account performance simulation, high accuracy of the geometric
specificities of the combustion process in Wankel engine. similarity at TDC is much more important then at BDC,
where combustion does not take place.
GEOMETRIC SIMILARITY
ALGORITHM
The main purpose of the algorithm is to ensure that the VP
engine would meet the criteria of geometric similarity listed
in the previous section. It means that the VP engine has to be
of the same displacement, compression ratio, behavior of
working chamber volume and surface-to-volume ratio change
versus shaft angle as the modeled Wankel engine.
DISCHARGE COEFFICIENTS FOR DC values were calculated for each studied position of the
rotor using the isentropic gas flow rate equation for subsonic
INTAKE AND EXHAUST PORTS velocity of outflow [20], where the real process effects were
Values of discharge coefficients (DC) for intake and exhaust taken into account to some extent by using a typical range of
ports are very important because they significantly affect the polytrophic exponent n values:
calculations of gas exchange in the engine working chamber.
It is clear that they greatly influence the accuracy of the
engine performance simulation. Therefore, DC data used as
the inputs in a simulation approach must be found as accurate
(2)
as possible.
Where: Q - gas mixture flow rate; DC - discharge coefficient;
Obviously, an accurate DC assessment would be based on
A - area of the port; n - polytropic exponent; ρo - upstream
experimental measurements, when air is blown through the
real orifice. Unfortunately, these results are not always mixture density; p0 - upstream pressure; p - downstream
available on early stages of engine development. Therefore a pressure.
method of flow simulation through the intake and exhaust
ports was developed. These calculations were performed for
few different positions of the rotor relative to the
corresponding port. Appropriate charts were prepared using
Solid Works software. Examples of this virtual blowing are
presented in Figures 7 (for intake port) and 8 (for exhaust
port).
Gas flow rate values through the port were computed using
the FLUENT software for each studied position of the rotor.
Unburned mixture with air/fuel ratio of 14.5 (typical for SI
Wankel engines) and burned mixture of N2/CO2/H2O vapor
(typical for complete combustion) were used for the
simulation of gas flow through the intake and exhaust port, Figure 8. Simulation scheme for exhaust port blowing
respectively.
Input data, such as pressure and temperature in the working
chamber, that are required for the calculation carrying out,
have been assessed using a typical indicatory diagram of a
Wankel engine. Upstream data for the exhaust port and
downstream data for the intake port were spatially averaged
for the combustion chamber volume for each considered
opening area of the corresponding port. Downstream data for
the exhaust port and upstream data for the intake port were
taken as ambient conditions (1 bar and 20 degrees of
Celsius). Since pressure values in the working chamber that
have been assessed based on the typical indicatory diagrams
cannot be sufficiently accurate, the calculations were made
for several values of the pressure in the vicinity of the
calculated point. This allowed assessment of DC sensitivity
Figure 7. Simulation scheme for intake port blowing. to pressure value p in the working chamber. Fig. 9 illustrates
example of calculation results for 50% opening of exhaust
Molecular weight of each gas mixture was calculated by
port. As can be seen from this Figure, minimal DC value was
equation:
found for the combination of the minimal pressure and
polytrophic exponent values. Accordingly, the maximal DC
(1) value anticipated for the highest values of pressure and n.
where: Ci - mole fraction of the component “i” Mi - Deviation of the computed DC values in the realistic range of
pressures and polytrophic exponents was found to be less
molecular weight of the component “i”.
than ±12%. In this work the DC values for each port opening
were assumed to be the average of those found by the
The data that have been used for the calculation of gas
computations for nine combinations of pressure value in the
mixtures molecular weight and calculation results for intake
working chamber and polytropic exponent in the range of
and exhaust ports are given in Table 3.
THIS DOCUMENT IS PROTECTED BY U.S. AND INTERNATIONAL COPYRIGHT.
It may not be reproduced, stored in a retrieval system, distributed or transmitted, in whole or in part, in any form or by any means.
Downloaded from SAE International by Leonid Tartakovsky, Tuesday, October 16, 2012 03:48:23 AM
1.5<p<2.5 bar (exhaust port), 0.7<p<0.9 bar (intake port) and Table 4. DC values calculated for Wankel engine and
1.2<n<1.4. those typical for reciprocating engines
were carried out for the same ambient and cooling conditions
that were observed under the engines tests. For each
calculated regime piston surface temperature of the virtual RP
engine was assumed equal to that one of the simulated
Wankel rotor face, temperatures of cylinder head and
cylinder surfaces were assumed equal to these of the Wankel
housing inner surface on the base of experimental data. HTC
values were computed as described in the previous section.
The engine tests and calculations were carried out over
regimes of full load and propeller curves. The latter was
needed since the simulated Wankel engines are intended for
UAV propulsion, i.e. they are loaded by propeller where the
relationship between the engine load P and speed n is
described by the equation P=kn3.
Figure 15. Wankel engines 802 and 802W: predicted
Figures 13,14,15 show satisfactory correspondence of the (lines) and measured (dots) values of the fuel
calculated and measured values of brake power and fuel consumption over the propeller curve.
consumption at various operation regimes of the 802W and
802 engines. As can be seen from Figures 13, 14, 15, the maximal
discrepancy of the measured and predicted results is about
5% and 3-10% for the brake power and fuel consumption,
respectively. This indicates the acceptability of the model
presented here.
simulation of a Wankel engine. The applied methodology of 5. Shih, T., Schock, H., and Ramos, J., “Fuel-Air Mixing and
the heat transfer coefficient computing allowed accounting Combustion in a Two-Dimensional Wankel Engine,” SAE
for such features of Wankel engine as “hot” and “cold” stator Technical Paper 870408, 1987, doi: 10.4271/870408.
zones separation, unfavorable shape of working chamber and
6. Lee, C. and Schock, H., “Regressed Relations for Forced
the charge rotational movement.
Convection Heat Transfer in a Direct Injection Stratified
Charge Rotary Engine,” SAE Technical Paper 880626, 1988,
For the case when discharge coefficients cannot be
doi: 10.4271/880626.
experimentally obtained (the common situation at early
engine development stages), a method of blowing simulation 7. Ramos, J.I., Shih, T.I-P., Schock, H.J., “Wankel engine
through the intake and exhaust ports was developed. This modeling”, Proceedings of the International Centre for Heat
method includes compilation of charts for intake and exhaust and Mass Transfer: 469-483, 1989.
port zones, when each of the charts has different rotor
8. Li, Z., Steinthorsson, E., Shih, T., and Nguyen, H.,
position relative to the port (i.e. different rate of the port
“Modelling and Simulation of Wankel Engine Flow Fields,”
opening). Unburned mixture with air/fuel ratio of 14.5 and
SAE Technical Paper 900029, 1990, doi: 10.4271/900029.
burned mixture of N2/CO2/vapor H2O were assumed for
blowing through the intake and exhaust port respectively. The 9. Bartrand, T.A., and Willis, E.A., “Performance of a
gas flow rate through the port can be computed using any Supercharged Direct-Injection Stratified-Charge Rotary
available CFD software for different positions of the rotor, Combustion Engine”, NASA technical memorandum 103105,
with appropriate calculation of discharge coefficients. 1990.
10. Kim, K.J., Chae, J.O., and Chung, T.Y., “A Performance
Combustion rate and heat release behavior were simulated by Simulation for Spark Ignition Wankel Rotary Engine,” SAE
using the conventional approach for piston engines - the Technical Paper 912479, 1991, doi: 10.4271/912479.
Wiebe equation. Values of the constants in the latter were
appropriately changed based on the available literature data 11. Raju, M., and Willis, E., “Three Dimensional Analysis
for Wankel engines. and Modeling of a Wankel Engine,” SAE Technical Paper
910701, 1991, doi: 10.4271/910701.
This approach was demonstrated through the example of the 12. Raju, M., “Heat Transfer and Performance
802 and 802W engines brake power and fuel consumption Characteristics of a Dual-Ignition Wankel Engine,” SAE
simulation at various operation regimes. A comparison of the Technical Paper 920303, 1992, doi: 10.4271/920303.
predicted Wankel engines performance with the measured
data has shown satisfactory correlation. 13. Bartrand, T.A., and Willis, E.A., “Rotary engine
performance computer program. User's guide”, NASA
The developed simulation approach can be used for Contractor Report 191192, October, 1993.
optimization of Wankel engine parameters, such as ignition 14. Norman, T.J., “A performance model of a spark ignition
advance map, intake and exhaust ports timing and geometry, Wankel engine: including the effects of crevice volumes, gas
air/fuel ratio, as well as assessment of the influence on engine leakage and heat transfer”, M.S. thesis, Massachusetts
performance of air filter and exhaust muffler characteristics, Institute of Technology, Massachusetts, June 1983.
flight altitude, etc.
15. Wendekerer, M., Grabovski, L., Pietrikovski, K. and
REFERENCES Magryta, P., “Phenomenological model of a Wankel engine”,
Silniki Spalinowe, 50(3), 2011.
1. Danieli, G., Keck, J., and Heywood, J., “Experimental and
16. Handschuh, R.F. and Owen, A.K., “Analysis of apex seal
Theoretical Analysis of Wankel Engine Performance,” SAE
friction power loss in rotary engines”, NASA/TM
Technical Paper 780416, 1978, doi: 10.4271/780416.
-2010-216353, September 2010.
2. Horsfield, B., “Computer Modeling [Wankel Rotary
17. Poojitganont, T., Izweik, H.T., and Berg, H.P., “The
Combustion Engine]” Electronics & Computing Monthly
simulation of Flow Inside the Wankel Combustion
3(6): 56-7, 1983.
Chamber”, http://www.tsme.org/ME_NETT/ME_NETT20/
3. Weston, K.C., “Computer Simulation of a Wankel Rotary article/pdf/tsf/TSF031.pdf, Nov. 2011.
Engine - Analysis and Graphics”, Proceedings of the 1986
18. Yamamoto, Kenichi, “Rotary engine”, Published by
Summer Computer Simulation Conference, USA, July 28-30,
Toyo Kogyo Co., LTD. 1969.
1986.
19. Automotive handbook, 8th edition, Robert Bosch GmbH,
4. Roberts, J., Norman, T., Ekchian, J., and Heywood, J.,
2011.
“Computer Models For Evaluating Premixed and Disc
Wankel Engine Performance,” SAE Technical Paper 860613, 20. Mechanical engineering handbook, V.2, Moscow, 1961
1986, doi: 10.4271/860613. (in Russian).
THIS DOCUMENT IS PROTECTED BY U.S. AND INTERNATIONAL COPYRIGHT.
It may not be reproduced, stored in a retrieval system, distributed or transmitted, in whole or in part, in any form or by any means.
Downloaded from SAE International by Leonid Tartakovsky, Tuesday, October 16, 2012 03:48:23 AM
21. http://www.gtisoft.com/applications/
a_Engine_Performance.php
22. Heywood, John B., “Internal combustion engine
fundamentals”, McGraw -Hill Inc., 930 p., 1988.
23. Nusselt, W., “Das Grundgesetz des Wärmeüberganges”,
Gesundheits-Ingenieur, 38 (42): 477-482 and (43): 490-496,
1915.
24. Nusselt, W., “Der Wärmeübergang im Rohr”, Z. VDI,
61, 1917.
CONTACT INFORMATION
Dr. Leonid Tartakovsky
Faculty of Mechanical Engineering
Technion - Israel Institute of Technology
Technion City, Haifa 32000, Israel
tartak@technion.ac.il
ACKNOWLEDGMENTS
The authors are grateful to MAFAT, Israeli Ministry of
Defense for the financial support of this project.
DEFINITIONS/ABBREVIATIONS
BDC - bottom dead center
DC - discharge coefficient
HTC - heat transfer coefficient
RP - reciprocating piston
SI - spark ignition
TDC - top dead center
UAV - unmanned aerial vehicle
VP - virtual piston
THIS DOCUMENT IS PROTECTED BY U.S. AND INTERNATIONAL COPYRIGHT.
It may not be reproduced, stored in a retrieval system, distributed or transmitted, in whole or in part, in any form or by any means.
Downloaded from SAE International by Leonid Tartakovsky, Tuesday, October 16, 2012 03:48:23 AM
APPENDIX
APPENDIX A
Rotor lobe surface area Ac is a constant value and was calculated by using the modified approximating formula from [14]:
(A1)
(A2)
(A3)
m is multiplier considering rotor lobe recess.
Following [14], in the equation (A3) values of R and e must be inserted in inches.
Instantaneous value of the housing surface area as function of the angle of rotor rotation was calculated analytically using the
following equation derived in this work:
(A4)
Instantaneous value of the side plate surface area consists of two parts - the first is the constant one Acompr (corresponds to
compression volume) and the second is the variable one Adispl (corresponds to displacement volume):
(A5)
Values of these parts were found using the following equations [16]:
(A6)
(A7)
APPENDIX B
THIS DOCUMENT IS PROTECTED BY U.S. AND INTERNATIONAL COPYRIGHT.
It may not be reproduced, stored in a retrieval system, distributed or transmitted, in whole or in part, in any form or by any means.
Downloaded from SAE International by Leonid Tartakovsky, Tuesday, October 16, 2012 03:48:23 AM
Fig. B1. Flow chart of virtual piston-to-Wankel engines geometric similarity algorithm. For designations see Fig.B2 and a list
below.
Symbols designation
Ac - rotor lobe area;
As - area between rotor lobe arc and chord XY;
Vcr - volume of rotor recession;
Δα - angle step of calculations;
THIS DOCUMENT IS PROTECTED BY U.S. AND INTERNATIONAL COPYRIGHT.
It may not be reproduced, stored in a retrieval system, distributed or transmitted, in whole or in part, in any form or by any means.
Downloaded from SAE International by Leonid Tartakovsky, Tuesday, October 16, 2012 03:48:23 AM
The Engineering Meetings Board has approved this paper for publication. It has Positions and opinions advanced in this paper are those of the author(s) and not
successfully completed SAE's peer review process under the supervision of the session necessarily those of SAE. The author is solely responsible for the content of the paper.
organizer. This process requires a minimum of three (3) reviews by industry experts. SAE Customer Service:
Tel: 877-606-7323 (inside USA and Canada)
All rights reserved. No part of this publication may be reproduced, stored in a
Tel: 724-776-4970 (outside USA)
retrieval system, or transmitted, in any form or by any means, electronic, mechanical, Fax: 724-776-0790
photocopying, recording, or otherwise, without the prior written permission of SAE. Email: CustomerService@sae.org
ISSN 0148-7191 SAE Web Address: http://www.sae.org
Printed in USA