Parametric Knocking Performance Investigation of Spark Ignition Natural Gas Engines and Dual Fuel Engines
Round 1
Reviewer 1 Report
All the comments (minor) from the previous review have been addressed.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
This is a very interesting study, providing the background for the proper usage of semi-empirical models for the optimum determination of engine operating parameters of NG engines. Moreover, the application of this model on two types of NG engines is useful for comparative studies. In general, this is a well-written paper, which is suggested to be published. It is believed that it would be for the benefit of the value of the paper if the authors consider the following comments and making the corresponding (if any) changes in the final version of the paper:
Line 16: “which sufficiently represents the in-cylinders processes providing adequate prediction of the in-cylinder temperature, which is crucial for the knocking prediction”. It is suggested to rephrase this sentence taking into account that with Wiebe function the combustion process is not represented but just the combustion rate is mathematically determined, as well as with a 2-zone model the prediction of in-cylinder temperature is not adequate, especially if someone needs to concentrate on local high temperature spots that are the source of knocking.
Line 28: “engines, thus it can provide useful support for designing future sustainable gas engines”. Please consider to better describe the actual scope of the work taking into account the inherent limitations of the sub-models used. It is suggested to focus on existing engines where the models used can be calibrated / validated.
Line 168 “As concluded from the discussed literature [2, 47-51], a two-zone 0-D model could be an effective tool for the knocking prediction in NG engines as it is capable of characterising the end-gas zone temperature with the simplest combustion zone division.”. It is suggested to be clear that auto-ignition occurs locally in places where high temperature spots exist and this cannot be predicted with a 2-zone model. In this respect, a 2-zone model can only be used to qualitatively take into account the effect of engine operating conditions and accordingly feed adequate empirical knock models.
Line 177 “which may exhibit variable”: It is suggested to correct with: different
Line 195 “which could form the basis for a more systematic optimal engine design studies in the future.”
Line 224 “…injection pulse. Thus, a fixed air-fuel equivalence ratio is used to characterize the NG injector in GT-Power.”: If this is the case, then there is no fuel injection model. Please clarify this.
Line 249, eq. 4: Please define Δφ
Line 257 “….The model can be developed into a condition-independent model if enough experimental data or CFD simulations are available to find the unique set (or sets)”. Please clarify the meaning of the term condition-independent. It may lead to misunderstandings.
Line 262, eq.6 : Since you have different duration of combustion for pilot and NG fuel, you should present the corresponding formula when pilot fuel has been burnt.
Line 264 “where, b1 represents the weight factor of the pilot diesel fuel combustion, 0< b1 <1.”: Please clarify how the b1 parameter is determined?
Line 268 “combustion profile for both the diesel and the gas modes, were calibrated at different loads (at steady”: Please clarify if these parameters have been calibrated only at different loads or also at different ignition timing, residual gas portion, rotational speed, to take into account the effect of parameters in eqs 3,4,5 ?
Line 271 the combustion model parameters are stored in a database as functions of the engine load and rotational speed.”: Please clarify if the calibration is also done at various engine speeds. If not please justify the methodology followed.
Line 279 Eq.7: Do you take into account fuel evaporation (pilot injection) ?
Line 281” is the cylinder pressure; V1 is the unburnt zone volume; Q1 is the unburnt zone heat transfer rate; mf is the fuel mass; hf is the fuel specific enthalpy; ma is the air mass; ha is the air specific enthalpy; mf,i is”: Please explain the difference between the fuel mass and the fuel-injected mass.
Line 284, eq 8: Please check the subscripts (2) and the expression. How do you take into account combustion?
Line 292 “(bar); cm is the average piston speed (m/s); C1 is the coefficient depending on the airflow velocity; C2 is the coefficient related to combustion chamber shape.”: How do you calibrate these parameters? Do you keep these values constant for both engine designs and at all operating conditions? Please clarify.
Line 308 “crank angle (θE)”: Please define θE
Line 347 “of the two developed models, respectively. However, the knocking prediction model was not validated by using experimental data as the engine operation with knocking conditions was not”: Please clarify how did you calibrated these parameters ?
Line 354, Table 1. “Ignition/Injection Timing °CA BTDC 36 5”: Please specify that this is the ignition timing for the SI engine and the pilot fuel injection timing
Line 356 “The experimentally measured parameters for the two investigated engines include in-cylinder pressure, heat release rate, fuel mass flow, inlet air mass flow, exhausted manifold pressure and emissions”, Please correct: heat release rate is not measured. Also, add mass flow rate when the flow is mentioned.
Line 357 “pressure, heat release rate, fuel mass flow, inlet air mass flow, exhausted manifold pressure and”. Please correct: exhaust. Moreover, do you consider also the inlet manifold pressure (in Diesel engine) ?
Line 371 “corresponding methane number obtained by using the Wärtsilä Methane Number Calculator [80] is 100,”. Please make a comment if the Methane Number is a parameter that affects the combustion mechanism and the values of the estimated parameters. If this is the case, how is this considered in the proposed model?
Line 391 “For the Engine A, the CNG is injected”: It is suggested to use the term NG
Line 392 “mixture that enters into the engine cylinders. The selected NG injection model keeps the air-fuel”. Probably there is no NG injection model. Please clarify if you simulate NG injection process or pre-determine the NG mass injected via the known A/F ratio.
Line 395 “analysing the measured HRR curve. The crank angle at the SOC was obtained by using the scheduled”. Does this timing coincide with the one obtained from HRR ? Why not using one reference source of data (HRR diagram) for the start and end of combustion?
Line 394 the 2135 engine, the Wiebe model parameters (SOC, Δφ and m) for 100% load were calibrated by analysing the measured HRR curve. The crank angle at the SOC was obtained by using the scheduled spark timing at different operating conditions, whilst the combustion duration was calculated by”, Do you calibrate also in other than 100% engine load conditions? Please make this clear.
Line 399 Then, the Wiebe model parameters of other operating conditions were calculated by employing Equations (3)-(5).”, Please specify which are these "other" operating conditions?
Line 401 “The developed model was validated against the measured in-cylinder pressure for the Engine”: The description is very brief although this is claimed to be one of the innovative features of this study. Before validating the model, you must determine the parameters based on (perhaps) a polynomial equation. Please describe in more detail the procedure followed as well as the final expressions derived for each parameter.
Line 402 “A at 25% and 50% loads. “ Are these operating conditions among the ones where the model parameters have been calibrated?
Line 411 “position is less than 1°CA. Thus, it can be inferred that the developed model is able to predict with adequate accuracy the performance of the Engine A”. It is suggested to validate the model at the full operating range of the engine (speed / load). This will provide a better justification of the claimed adequate prediction of engine performance
Line 413, Table 4. Please add also the parameters from eq. 3-5
Line 427 “The HRR obtained by the KiBox combustion analyser [74] is calculated from the measured pressure diagram by assuming that natural gas and diesel fuel burn proportionately according to their supplied mass ratio”: The HRR provides the rate of combustion at each time step. Could you please clarify if the proportionality factor used is determined at each time step according to the instantaneous mass flow rate of each fuel (which would need a corresponding model for the fuel injection process), or this is determined based on the total fuel mass injected?
Line 431 “The Weight Factor b1, which is used to specify the energy percentages of pilot fuel, can be acquired by calculating the energy proportion by using the mass flow and the LHV of diesel fuel.”. Please provide more details on how this factor is calculated and what are the assumptions made.
Line 435 “The SOC was obtained from the measured cylinder pressure diagrams (abrupt rising of pressure in comparison with the motoring cylinder pressure).”: Since you have the HRR why don't you rely on it? Please make a comment.
Line 439 for the natural gas. The remaining three Wiebe parameters, namely the combustion duration, the Wiebe exponent for diesel fuel and the Wiebe exponent of natural gas, can be obtained by a curve fitting method.”: Since you have simultaneous combustion of two fuels with different combustion rates how can you determine a singular set of values for the unknown parameters based on one overall HRR rate diagram? The problem seems to be under-defined. Please explain in more detail the calibration procedure followed.
Line 443 “With the calibrated Wiebe combustion model parameters at the 4 operating points, an interpolation model was employed to predict the Wiebe combustion parameters at the other operating points”, Please provide the formula obtained for each of the calibration parameter defined.
Line 458 Table 6. “Wiebe parameters for HRR calculation in the YC6K duel-fuel engine”. It is suggested to have uniform reference to engines.
Line 459 “Figure 4. In-cylinder pressure comparison of Engine B at different operation conditions: (a) 50% load; (b) 85% load.”, It is suggested to limit the presented range of CA degs inside the close engine cycle (which is actually simulated)
Line 482 “The spark timing of 2135 engine is advanced by 5 degree at the 100% load operating condition, whilst the pilot injection timing of YC6K engine is advanced by 5 degree.”: Are these operating conditions outside the range where the model parameters have been calibrated? Wiebe functions are not adequate for off-design simulation. Please make a comment. Moreover, how is this value of start of ignition or start of injection is determined to be at the knocking boundaries of both the engines examined?
Line 495 “equivalence ratio and the ignition timing variations within specific ranges around their baseline”: Please clarify how these ranges have been determined?
Line 500 “All the parameters except for the compression ratio remain the same with those at the engine nominal conditions.” Do you use the same parameters for the Wiebe functions? Could you please justify that the Wiebe parameters are not affected by the CR and A/F ratio? (i.e. parameters as: duration of combustion, shape of HRR curve, etc.)
Line 520 “For a specific compression ratio, the knocking IMEP is obtained by increasing the injected natural gas mass while keeping a fixed air-fuel equivalence ratio till the knocking occurs”: Therefore this is not the effect of Air/Fuel equivalence ratio, but the effect of increasing boost pressure (which is not possible for the naturally aspirated engine). Please make this clearer.
Line 553 “Because the SOC can be apparently obtained by analysing the measured in-cylinder pressure, the combustion model of Engine B is developed by using the obtained SOC instead of employing the injection timing and the estimated ignition delay”: Please clarify. This is a theoretical investigation. Do you have experimental cylinder pressure from which you estimate the SOC ? Why don't you need to simulate the ignition delay?
Line 563 “all the parameters except for the compression ratio are kept the same with those at the nominal condition.”: Does this mean that you keep the same also the inlet and exhaust pressure (boundary conditions) ? Please justify if this is the case.
Line 580 “denotes a knocking occurrence and suggests that the air-fuel equivalence ratio should be reduced from original 2.1 to below 2.”: It seems strange that the engine B has nominal operating conditions inside the high risk knocking region. It has been mentioned that no experimental data are available to calibrate the knock model. Please check and make the situation clearer.
Line 605 “In order to minimise the influence of the combustion model parameters”: Please correct. The actual scope should be to maximize the influence, for the correct prediction of the HRR.
Line 606 “the Wiebe functions parameters for the pilot diesel fuel and natural gas are calibrated for six engine operating points, so that the engine cylinder pressure diagrams match the measured ones with adequate accuracy. The employed Wiebe parameters for the Engine B are shown in Table 10.”: This description belongs to the model calibration description. It is suggested to transfer it with Table 10 in the corresponding section.
Line 610 “Subsequently, the energy proportion of the pilot diesel fuel (b1 as in eq. 6) varied, whilst the Wiebe functions parameters of both diesel and natural gas were kept the same with those provided in”: Based on values presented this is kept constant and equal to 10% which in general seems to be a high value. Please check and comment on this.
Line 660 “range when the diesel energy proportion increases from 5% to 15%, which is attributed to a faster flame propagation caused by more ignition points.”: This is an assumed physical explanation that cannot be captured with the model used. The real cause of the observed behavior is that keeping the b1 constant while increasing the energy proportion of the pilot fuel (keeping also the predetermined combustion duration constant), more fuel is burnt (mathematically imposed) at the initial stages of combustion. Please make this clear.
Line 662 “advanced with a stronger knocking intensity for the cases where the pilot fuel mass increases keeping a constant fuel energy input.” Please make clear that the total energy input is kept constant but indirectly you alter the rate of the energy input.
Line 667 “The parametric investigation provides suggestions for engine design settings (compression”: This is a general comment. It is suggested to make clear that the validity of the methodology presented is directly dependent on the existence of adequate experimental data in the operating range of the parameters examined. This is due to the absence of a simulation model related to the physical and chemical processes that affect combustion. Although (in my opinion) this type of model might not be adequate for determining the engine DESIGN setting (i.e. used for the design of a NEW engine / off-design investigation), it is of great value when used to define the optimum settings of existing engine parameters (inside the design space where the model parameters have been calibrated), offering significant advantages compared to conventional physical models, and being ideal for digital twin applications and/or safety assessment, real-time condition monitoring, etc. It is suggested to make this clear throughout the text and in the conclusions if this is in accordance to the scope of your study.
Line 677 “The models were first validated against experimental data and subsequently were used for the parametric investigation to”: It is suggested to make clear that no experimental data were available for the knock model.
Line 690 “ratio for the Engine B is suggested to be reduced below 2, however experimental verification for this limit is required.”: Since this is an operating point where experimental data are available, please make clear what are the missing data?
Line 698-707: “The following recommendations were provided and can be used during the design and operation stages for SI natural gas engines and DF engines. Prior to attempting to improve the performance of an engine (SI NG or DF) by increasing the compression ratio or advancing ignition timing, knocking ehaviour analysis must be performed by considering the knocking occurrence, onset position and intensity as well as the cylinder peak pressure upper limit. The ignition timing of the SI natural gas engines and DF engines must be controlled precisely after the knocking limit in order to suppress knocking occurrence. When operating at high load conditions, the pilot fuel mass of DF engine is suggested to remain at a low level to avoid deteriorating the knocking behaviour.”: These are general rules for engine design, which are not the direct outcome (conclusions) of the present work. It is suggested to be omitted.
Line 732: There are some missing information (mainly Journal name) in the following references: 2, 14, 16, 29, 62, 63, 83, 86. Please correct.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
This paper presents results of 1D simulations on CI and SI NG engines
A few comments:
- The formatting of the paper is really bad - this needs to be fixed.
- Did you consider running the engines at fixed IMEP instead? This is more likely to be the way the engines are used in the "real world" where there is a power demand
- Did you do experimental repeats? Can you quantify the uncertainty in the 1-D model? Fig 4 onwards would be much better if they had error bars on them
- Table 7 and Fig 4 - is this a single cycle match or the average of a number of cycles?
- How was the results in Table 3 measured? This should be stated
- Throughout there is confusion between Engine A&B and 2135/YC6K - could this be clarified?
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
