Knocking prediction method, knocking prediction device, combustion device design method, power generation device design method, vehicle design method, and fuel design method

Abstract

The purpose of the present disclosure is to provide a knocking prediction method capable of further improving accuracy.　The present disclosure pertains to a knocking prediction method for predicting knocking that may occur in a premixed combustion field, and adopts a means of predicting knocking on the basis of the knowledge that, in a premixed combustion field, a temporal change in a normalized fuel mass fraction and a normalized temperature in a zero-dimensional homogeneous ignition is equivalent to a spatial change in a one-dimensional laminar flow premixed flame.

Description

Knocking prediction method, knocking prediction device, combustion device design method, power generation device design method, vehicle design method, and fuel design method

[0001] The present disclosure relates to a knocking prediction method, a knocking prediction device, a combustion device design method, a power generation device design method, a vehicle design method, and a fuel design method.

[0002] As is well known, in internal combustion engines that use premixed combustion, such as gasoline engines, the prevention of knocking is an important design issue. Knocking is a phenomenon accompanied by a shock wave generated by the spontaneous ignition of unburned gas in the combustion chamber, and can potentially damage the engine. In internal combustion engine design, the pressure and temperature (compression ratio) of the combustion chamber are reduced to prevent knocking from occurring.

[0003] Meanwhile, the "Livengood-Wu integral" is a well-known knocking prediction method. This "Livengood-Wu integral" is a method of estimating the ignition delay time when the pressure and temperature in the combustion chamber are changed and integrating it over time, and when the time integral value reaches "1", auto-ignition, i.e., knocking, occurs. The "Livengood-Wu integral" is widely used in the design of actual engines, as shown in Patent Documents 1 and 2 below.

[0004] Japanese Patent Publication No. 2004-332584 Japanese Patent Publication No. 2007-127004

[0005] The Livengood-Wu integral is a simple method that can predict the occurrence of knocking (auto-ignition) to a certain extent, but its prediction accuracy is not necessarily sufficient. In order to make the most of the capabilities of an internal combustion engine functioning as a power generation device, it is highly desirable to establish a knocking prediction method with higher prediction accuracy than the Livengood-Wu integral.

[0006] The present disclosure has been made in consideration of the above-mentioned circumstances, and aims to provide a knocking prediction method and a knocking prediction device that can further improve accuracy. The present disclosure also aims to provide a combustion device design method, a power generation device design method, a vehicle design method, and a fuel design method that can improve the substantial performance of a combustion device, a power generation device, a vehicle design method, and a fuel design method.

[0007] A knocking prediction method according to a first aspect of the present disclosure is a knocking prediction method for predicting knocking that may occur in a premixed combustion field, and predicts knocking based on the finding that in the premixed combustion field, the temporal changes in normalized fuel mass fraction and normalized temperature in zero-dimensional homogeneous ignition are equivalent to the spatial changes in a one-dimensional laminar premixed flame.

[0008] A knocking prediction method according to a second aspect of the present disclosure is a knocking prediction method for predicting knocking that may occur in a premixed combustion field, and in a characteristic diagram showing the relationship between normalized temperature and normalized fuel mass fraction of fuel, a fuel having a Lewis number greater than 1 is deemed to have a possibility of knocking, and a fuel having a Lewis number smaller than 1 is deemed to have no possibility of knocking.

[0009] A knocking prediction method according to a third aspect of the present disclosure is a knocking prediction method for predicting knocking that may occur in a premixed combustion field, and includes a calculation step of sequentially calculating the combustion velocity when the inlet temperature is gradually increased by one-dimensional steady-state analysis of the premixed combustion field, a determination step of determining whether the combustion velocity has become impossible to obtain because a flame can no longer exist in the premixed combustion field, and a generation condition acquisition step of acquiring the inlet temperature corresponding to the last acquired combustion velocity as the generation condition for the knocking.

[0010] A knocking prediction method according to a fourth aspect of the present disclosure is the knocking prediction method according to the third aspect, wherein in the calculation step, the size of the region including the unburned mixture, the flame reaction zone, and the burned gas in the premixed combustion field is calculated by reducing the size to the limit that can be encompassed by the flame.

[0011] A knocking prediction method according to a fifth aspect of the present disclosure is the knocking prediction method according to the third or fourth aspect, in which a fuel having a Lewis number greater than 1 is targeted for prediction.

[0012] A knocking prediction device according to a first aspect of the present disclosure is a knocking prediction device that predicts knocking that may occur in a premixed combustion field, and includes a calculation means that sequentially calculates the combustion velocity when the inlet temperature is gradually increased by one-dimensional steady-state analysis of the premixed combustion field, a determination means that determines whether the combustion velocity has become impossible to obtain because a flame can no longer exist in the premixed combustion field, and an occurrence condition acquisition means that acquires the inlet temperature corresponding to the last acquired combustion velocity as the occurrence condition for the knocking.

[0013] A knocking prediction device according to a second aspect of the present disclosure is the knocking prediction device according to the first aspect, wherein the calculation means calculates the size of the region including the unburned mixture, the flame reaction zone, and the burned gas in the premixed combustion field by reducing it to the limit that can be contained by the flame.

[0014] A knocking prediction device according to a third aspect of the present disclosure is the knocking prediction device according to the first or second aspect, wherein the calculation means performs calculations on fuels having a Lewis number greater than one.

[0015] A combustion device design method according to the first aspect of the present disclosure designs a combustion device based on knocking occurrence conditions estimated based on the knocking prediction method according to the third or fourth aspect.

[0016] The method for designing a power generating device according to the first aspect of the present disclosure designs a power generating device based on knocking occurrence conditions estimated based on the knocking prediction method according to the third or fourth aspect.

[0017] A vehicle design method according to a first aspect of the present disclosure designs a vehicle based on knocking occurrence conditions estimated based on the knocking prediction method according to the third or fourth aspect.

[0018] The fuel design method according to the first aspect of the present disclosure designs fuel based on knocking occurrence conditions estimated based on the knocking prediction method according to the third or fourth aspect.

[0019] According to the present disclosure, it is possible to provide a knocking prediction method, a knocking prediction device, a combustion device design method, a power generation device design method, a vehicle design method, and a fuel design method that are capable of further improving accuracy. Also, according to the present disclosure, it is possible to provide a combustion device design method, a power generation device design method, a vehicle design method, and a fuel design method that are capable of improving the substantial performance of a combustion device, a power generation device, a vehicle design method, and a fuel.

[0020] FIG. 1 is a flowchart showing a knocking prediction method according to an embodiment of the present disclosure. FIG. 2 is a block diagram showing a functional configuration of a knocking prediction device according to an embodiment of the present disclosure. FIG. 3 is a table showing various physical quantities handled in the theoretical explanation of an embodiment of the present disclosure. FIG. 4 is a first characteristic diagram showing knocking prediction characteristics according to an embodiment of the present disclosure. FIG. 5 is a second characteristic diagram showing knocking prediction characteristics according to an embodiment of the present disclosure. FIG. 6 is a third characteristic diagram showing knocking prediction characteristics according to an embodiment of the present disclosure. FIG. 7 is a characteristic diagram showing a comparative example of a knocking prediction method according to an embodiment of the present disclosure.

[0021] Hereinafter, an embodiment of the present disclosure will be described with reference to the drawings. Fig. 1 is a flowchart showing a knocking prediction method according to this embodiment. This flowchart shows a procedure for acquiring a flame propagation limit temperature, which is a condition for the occurrence of knocking, i.e., an initial temperature at which knocking occurs, for a premixed combustion field (combustion chamber) in which a premixed mixture of fuel and air (oxidizer) is ignited and burned by a forced means such as spark ignition.

[0022] This flowchart also shows an outline of the processing steps of a knocking prediction program (simulation program) executed by a predetermined computer. That is, the knocking prediction method according to this embodiment is realized by information processing (simulation) using a computer.

[0023] A computer that executes the knocking prediction method according to this embodiment, i.e., a computer that has the above-mentioned knocking prediction program installed therein, is the knocking prediction device according to this embodiment. That is, the knocking prediction device according to this embodiment is a device that outputs the conditions for knocking occurrence in a premixed combustion field by executing the knocking prediction program.

[0024] 2 is a block diagram showing the functional configuration of the knocking prediction device A according to this embodiment. As shown in FIG. 2, the knocking prediction device A includes at least a storage unit 1 that stores a knocking prediction program in advance, a calculation unit 2 that executes the knocking prediction program, an operation unit 3 that inputs an operator's operation instruction to the calculation unit 2, and an output unit 4 that outputs the calculation result of the calculation unit 2, i.e., the flame propagation limit temperature (initial temperature), to the outside.

[0025] Among the components of the knocking prediction device A, the memory unit 1, the calculation unit 2, and the operation unit 3 constitute the calculation means in this disclosure. As will be described in detail later, the memory unit 1, the calculation unit 2, and the operation unit 3 work together to sequentially calculate the combustion velocity when the inlet temperature is gradually increased by one-dimensional steady-state analysis of the premixed combustion field.

[0026] The memory unit 1, the calculation unit 2, and the operation unit 3 in the knocking prediction device A constitute the determination means of the present disclosure. That is, the memory unit 1, the calculation unit 2, and the operation unit 3 cooperate to determine whether or not the combustion velocity cannot be obtained because a flame cannot exist in the premixed combustion field.

[0027] Furthermore, the memory unit 1, the calculation unit 2, and the operation unit 3 in the knocking prediction device A constitute the occurrence condition acquisition means of the present disclosure. That is, the memory unit 1, the calculation unit 2, and the operation unit 3 cooperate to acquire the inlet temperature (flame propagation limit temperature) corresponding to the last acquired combustion speed as the knocking occurrence condition.

[0028] Here, the calculation means of the knocking prediction device A according to this embodiment corresponds to the calculation step in the knocking prediction method according to this embodiment. Also, the determination means of the knocking prediction device A according to this embodiment corresponds to the determination step in the knocking prediction method according to this embodiment. Furthermore, the raw condition acquisition means of the knocking prediction device A according to this embodiment corresponds to the occurrence condition acquisition step in the knocking prediction method according to this embodiment.

[0029] The premixed combustion field in this embodiment corresponds to the combustion chamber of an internal combustion engine (prime mover), such as a gasoline engine installed in a vehicle as a power generating device. In designing gasoline engines, measures such as lowering the compression ratio are taken to prevent knocking, i.e., to prevent self-ignition of unburned gas. Note that the above internal combustion engine includes a direct injection engine and an engine with a pre-combustion chamber.

[0030] That is, the knocking prediction method according to this embodiment relates to a design method for a combustion device equipped with a premixed combustion field (combustion chamber), a power generating device (prime mover), a vehicle, etc. The knocking prediction method according to this embodiment also relates to a design method for a fuel that burns in a premixed combustion field.

[0031] The knocking prediction method according to this embodiment is based on the premise that when a spatiotemporal transformation of a flame is applied to a premixed combustion field, the temporal changes in the normalized fuel mass fraction and normalized temperature in zero-dimensional homogeneous ignition are equivalent to the spatial changes in a one-dimensional laminar premixed flame.

[0032] That is, this knock prediction method is based on the theoretical basis of the equivalence between the time variation of normalized fuel mass fraction and normalized temperature in zero-dimensional homogeneous ignition and the spatial variation in one-dimensional laminar premixed flame.

[0033] According to this theoretical basis, the rate of decrease in the normalized fuel mass fraction in the preheating region of a premixed combustion field increases in the order of fuel Lewis number less than 1, Lewis number 1 (= ignition), and Lewis number greater than 1. Furthermore, it is derived that autoignition in the preheating region near the explosive transition can occur only for one-dimensional laminar premixed flames with Lewis number greater than 1.

[0034] In other words, the time change in flame propagation is comparable to that of a compressed premixed fuel or shock wave in a premixed combustion field. Furthermore, the Lewis number of the fuel in a premixed combustion field is a physical quantity that determines the existence of a premixed flame structure.

[0035] The theoretical basis of the knocking prediction method according to this embodiment will be described below. Note that Fig. 3 is a table listing various physical quantities that will be used in the following theoretical explanation.

[0036] The following equations (1) and (2) are the governing equations for zero-dimensional homogeneous ignition using normalized fuel mass fraction and temperature. These equations (1) and (2) are conservation laws of energy and chemical species under constant pressure and constant enthalpy, and are given by the mass fraction of the kth species, the chemical production rate of the kth species, the molecular weight of the kth species, the mass density, the temperature, the total number of chemical species, the enthalpy of the kth species, the average specific heat, and time.

[0037]

[0038] The mass density is given as the gas equation of state used for the interconversion of pressure, density, and temperature, as shown in the following equation (3): R in this equation of state (3) is a universal gas constant.

[0039]

[0040] Here, we consider the Legendre transformation from the multivariate function f. The total derivative of the multivariate function f, which indicates the characteristics of a gas mixture with temperature, mass density, mass fraction, and time as variables, is given by the following equation (4).

[0041]

[0042] In this equation (4), if the partial differential of the multivariate function f with respect to temperature is represented by "a", the multivariate function g shown in the following equation (5) is obtained. Furthermore, the total differential of this multivariate function g is expressed as the following equation (6), with the above-mentioned a, mass density, mass fraction, and time as variables. Based on equation (6), the total differential of the multivariate function g is expressed as the following equation (7).

[0043]

[0044] Furthermore, equation (8) is obtained based on a comparison between equations (6) and (7). Furthermore, since equation (5) can be defined as shown in equation (9), the relationship shown in equation (10) holds between multivariate function f and multivariate function g.

[0045]

[0046] Here, a candidate for the multivariate function f in Equation (8) and Equation (10) is the normalized temperature, and a candidate for the multivariate function g is the normalized fuel mass fraction in the case of the single-step chemical reaction model in Equation (8) or the normalized progress variable in the case of the multi-step chemical reaction model in Equation (10).

[0047] If it is assumed that once the multivariate function f or the multivariate function g at a certain time is determined, all of the variables (time, mass density, and fuel mass fraction) that represent the characteristics of the premixed gas in the premixed combustion field can be determined, that is, if it is assumed that the multivariate function f and the multivariate function g depend only on the time direction, then the above equations (8) and (10) can be rewritten from partial differential equations to ordinary differential equations.

[0048] As a result, when the normalized temperature is applied as a multivariate function f and the normalized fuel mass fraction is applied as a multivariate function g, the above equation (8) is expressed as the following equation (11), which shows the relationship between the time change of the normalized temperature and the time change of the normalized fuel mass fraction. Note that the normalized temperature is defined by the following equation (12), which uses the initial temperature and the final temperature as parameters. The normalized fuel mass fraction is defined by the following equation (13), which uses the initial fuel mass fraction and the final fuel mass fraction as parameters.

[0049]

[0050] Furthermore, when the normalized temperature is applied as a multivariate function f and the normalized progress variable of the multi-step chemical reaction model is applied as a multivariate function g, the above equation (10) can be expressed as the following equation (14), which shows the relationship between the time change of the normalized temperature and the time change of the normalized progress variable.

[0051]

[0052] Here, to enable the Legendre transformation, the normalized temperature, normalized fuel mass fraction, and normalized progress variable must be convex functions, and the total derivative of the normalized temperature must be greater than 0, the total derivative of the normalized fuel mass fraction must be less than 0, and the total derivative of the normalized progress variable must be greater than 0. This constraint always applies to single-step chemical reaction models, and typically also applies to multi-step chemical reaction models.

[0053] In the following description, for simplicity, the normalized fuel mass fraction is defined as the multivariate function g, and the signs of the normalized progress variables are simply inverted. Once either the normalized temperature or the normalized fuel mass fraction is determined, all remaining variables are determined. Based on this premise, equations (1) and (2) can be rewritten as the following equations (15) and (16) using the normalized temperature and normalized fuel mass fraction.

[0054]

[0055] As shown in equation (11) above, the relationship between normalized temperature and normalized fuel mass fraction is independent of time. Therefore, equation (11) can be rewritten using residence time as follows: Residence time is defined by the fluid parcel velocity and fluid parcel position as follows: Residence time is the total time spent by a fluid parcel within the control volume. The total derivative of such residence time is given by equation (19):

[0056]

[0057] By transforming equations (15) and (16) using equation (19), the following equations (20) and (21) are obtained. Furthermore, equation (17) can be rewritten as equation (22) below by using the parcel position. Furthermore, the relationship shown in equation (23) holds based on equation (22).

[0058]

[0059] On the other hand, the conservation equations of a one-dimensional laminar premixed flame are given by the following equations (24) to (26) consisting of the laminar burning velocity, the mixing diffusion coefficient of the kth species, and the thermal conductivity of the premixed gas. Furthermore, equations (25) and (26) can be rewritten as the following equations (27) and (28) using the normalized temperature and normalized fuel mass fraction.

[0060]

[0061]

[0062] Substituting equations (22) and (23) into equation (27) gives the following equation (29): As can be seen by comparing equation (29) with equation (28), the only difference between the two equations is the first term on the right-hand side.

[0063]

[0064] Here, the Lewis number of the fuel is defined by the following equation (30). Comparing equations (28) and (29) with this Lewis number in mind, it can be seen that both equations are equivalent when the Lewis number is 1. In other words, for a fuel with a Lewis number of 1, the temporal changes in the normalized fuel mass fraction and normalized temperature in zero-dimensional homogeneous ignition are equivalent to the spatial changes in one-dimensional laminar premixed flames.

[0065]

[0066] Figure 4 shows the relationship between zero-dimensional homogeneous ignition and one-dimensional laminar premixed flames in premixed combustion fields using hydrogen, methane, propane, and SNIa as fuels. SNIa represents the carbon state in the nuclear combustion process of Type Ia supernovae and has a very high Lewis number. Hydrogen, methane, and propane have Lewis numbers of 0.36, 0-96, and 1.95, respectively.

[0067] As mentioned above, for fuels with a Lewis number of 1, zero-dimensional homogeneous ignition, as indicated by the normalized fuel mass fraction and normalized temperature, is equivalent to one-dimensional laminar premixed flame, as indicated by the normalized fuel mass fraction and normalized temperature. Also, for fuels with a Lewis number less than 1, the one-dimensional laminar premixed flame profile is downwardly convex compared to the profile for fuels with a Lewis number of 1 (zero-dimensional homogeneous ignition).

[0068] On the other hand, for fuels with Lewis numbers greater than 1, the profile is convex above the profile for fuels with Lewis numbers of 1 (zero-dimensional homogeneous ignition) and may intersect with the profile for fuels with Lewis numbers of 1 (zero-dimensional homogeneous ignition) in the high-temperature region (reaction zone) of the flame where heat release is high.

[0069] This indicates that for fuels with Lewis numbers less than 1, the mass diffusivity of the fuel is greater than the thermal diffusivity, so the fuel diffuses preferentially rather than being heated. Conversely, for fuels with Lewis numbers greater than 1, the thermal diffusivity of the fuel is greater than the mass diffusivity, so the fuel heats preferentially rather than being diffused.

[0070] That is, the fact that fuels with Lewis numbers greater than 1 are heated preferentially rather than diffused is consistent with well-known technical knowledge. The effect of the Lewis number on the relationship between temperature and fuel mass fraction is theoretically discussed in the well-known book by Buckmaster, and the characteristics in Figure 4 are consistent with that theory.

[0071] In the following, we will consider the relatively low temperature region (preheating region) of the premixed combustion field where the influence of the Lewis number is evident. In this study, we will assume that the well-known Soret effect and Dufur effect can be ignored.

[0072] 4, the consumption of normalized fuel mass fraction at a given normalized temperature can be considered as the progress of the combustion reaction. That is, the progress of the combustion reaction is in the order of fuels with Lewis numbers less than 1, fuels with Lewis numbers of 1 (zero-dimensional homogeneous ignition), and fuels with Lewis numbers greater than 1.

[0073] For fuels with a Lewis number less than 1, autoignition (knocking) does not occur in the preheating region. That is, even if the initial temperature and pressure in the premixed combustion field are very high, a one-dimensional laminar premixed flame can exist. In contrast, for fuels with a Lewis number greater than 1, autoignition (knocking) may occur in the preheating region. That is, depending on the initial temperature and pressure in the premixed combustion field, a one-dimensional laminar premixed flame cannot exist.

[0074] For example, gasoline, one of the fuels, has a Lewis number greater than 1, so it is expected that autoignition (knocking) will occur whenever the unburned gas region rises to a temperature and pressure that makes flame propagation impossible. In other words, the characteristic time of the flame behind the shock wave is comparable to that of ignition, which can affect the transition from flame propagation to detonation. Also, since gasoline has a Lewis number greater than 1, there may be a region behind the shock wave where flame propagation is impossible.

[0075] Simulations (one-dimensional steady-state analysis) were performed on zero-dimensional homogeneous ignition and one-dimensional laminar premixed flames under the conditions of constant pressure and constant enthalpy for two types of premixed gases: a premixed gas of hydrogen and air with a Lewis number smaller than 1 (hydrogen fuel) and a premixed gas of n-heptane and air with a Lewis number larger than 1 (n-heptane fuel).

[0076] In this simulation, we performed a one-dimensional steady-state analysis of zero-dimensional homogeneous ignition and one-dimensional laminar premixed flames using a well-known multi-step chemical reaction model. The multi-step chemical reaction model for hydrogen fuel is the well-known UT-JAXA model, and the multi-step chemical reaction model for n-heptane fuel is the well-known Reduced SIP model.

[0077] In this simulation, the normalized temperature and normalized fuel mass fraction must be convex functions to maintain the Legendre transformation, so the calculation domain must be as short as possible. Therefore, in this simulation, the size of the domain including the unburned mixture, the flame reaction zone, and the burned gas in the premixed combustion field is set to the smallest possible size that can encompass the flame. Note that the numerical conditions (initial conditions) for the pressure and equivalence ratio are fixed values, for example, 0.1 MPa and 1.0.

[0078] In this simulation, a one-dimensional steady-state analysis of a one-dimensional laminar premixed flame was performed by varying the inlet temperature from 300 to 3000 K in increments of 100 K. When calculation failure was confirmed even when the calculation domain was reduced, numerical calculations were performed by varying the inlet temperature in increments of 1 K.

[0079] Furthermore, in this simulation, a one-dimensional steady-state analysis of a one-dimensional laminar premixed flame was performed, and then a numerical calculation of zero-dimensional homogeneous ignition was performed using an initial temperature that allows the calculation of a one-dimensional laminar premixed flame.

[0080] Such a simulation (one-dimensional steady-state analysis) corresponds to the knocking prediction method according to this embodiment. This simulation is the processing operation of the knocking prediction device according to this embodiment, i.e., information processing executed by the calculation unit 2 based on the knocking prediction program stored in advance in the storage unit 1.

[0081] That is, in the knocking prediction method according to this embodiment, an operator first operates the operation unit 3 to input the initial conditions required for the simulation, such as temperature, pressure, and the type and equivalence ratio of fuel, such as hydrogen fuel or n-heptane fuel, into the calculation unit 2 (step S1).

[0082] The calculation unit 2 uses the temperature, pressure, fuel type, and equivalence ratio input in step S1 and gradually increases the inlet temperature (calculation start temperature) to perform a one-dimensional steady-state analysis (calculation step) to calculate the flame propagation limit temperature for each fuel (step S2).

[0083] This one-dimensional steady-state analysis is a convergence calculation that shrinks the flame front and flame rear until the one-dimensional premixed flame structure becomes the narrowest calculation region that provides a unique solution. That is, the calculation unit 2 sequentially calculates the burning velocity when the inlet temperature is gradually increased using one-dimensional steady-state analysis of the premixed combustion field, but the calculated burning velocity may converge or diverge depending on the set value of the inlet temperature.

[0084] The calculation unit 2 performs a convergence determination (determination step) for each calculation of the burning rate for a certain inlet temperature (calculation start temperature) (step S3). That is, in step S3, the calculation unit 2 determines whether the burning rate has become unobtainable because a flame can no longer exist in the premixed combustion field. If the determination result in step S3 is "convergence," the calculation unit 2 sets the inlet temperature even higher and performs the burning rate calculation.

[0085] On the other hand, if the determination result in step S3 is "divergence," the calculation unit 2 sets the inlet temperature corresponding to the previous combustion speed calculation as the flame propagation limit temperature (step S4: occurrence condition acquisition step). That is, the calculation unit 2 acquires the flame propagation limit temperature (inlet temperature) corresponding to the last acquired combustion speed as the knocking occurrence condition. Then, the calculation unit 2 outputs the simulation result including the acquired flame propagation limit temperature as the knocking occurrence condition to the output unit 4.

[0086] Fig. 5 shows an example of the simulation results (one-dimensional steady-state analysis results), that is, calculation results including the knocking occurrence conditions output from the calculation unit 2 to the output unit 4. Fig. 6 shows an example of the relationship between the inlet temperature and the combustion velocity in the premixed combustion field based on such simulation results.

[0087] Figure 6 shows that for hydrogen fuel, the burning velocity increases with increasing inlet or initial temperature, and a propagating flame exists even at an inlet temperature of 3000 K. Figure 6 also shows that for n-heptane fuel, the burning velocity increases with increasing inlet temperature, but a propagating flame cannot exist above 1270 K.

[0088] In other words, the theoretical basis of this embodiment is that it theoretically shows that the relationship between the normalized fuel mass fraction and the normalized temperature is equivalent for zero-dimensional homogeneous ignition and one-dimensional laminar premixed flames, both of which have a Lewis number of 1 after space-time transformation, and that the Lewis number is a physical quantity that determines the existence of a premixed flame structure near the explosive transition of the flame in a premixed combustion field.

[0089] For 1D laminar premixed flames of fuels with Lewis numbers less than 1, ignition does not occur in the preheated zone, and therefore a flame structure always exists. In contrast, for 1D laminar premixed flames of fuels with Lewis numbers greater than 1, ignition may occur in the preheated zone, and no flame structure exists if the temperature is above a certain threshold.

[0090] In other words, if the characteristic time of ignition and flame is on the same order of magnitude as behind a shock wave or the explosive transition of an SI engine, then ignition and flame should be discussed via the Lewis number.

[0091] 7 shows a comparative example of this embodiment. That is, this comparative example is a calculation result of the "Livengood-Wu integral" by direct numerical simulation (DNS). It is well known that direct numerical simulation (DNS) accurately simulates the combustion state of a premixed combustion field. As is well known, the "Livengood-Wu integral" indicates that auto-ignition (knocking) occurs when the time integral value becomes 1.

[0092] However, according to this calculation result, knocking does not occur at the point in time (CFI) when the time integral value becomes 1. According to this calculation result, knocking occurs at the point in time (KO) after the time integral value exceeds 1. Therefore, the knocking prediction method based on the "Livengood-Wu integral" has problems in terms of accuracy.

[0093] In contrast, according to this embodiment, the accuracy is higher than that of the "Livengood-Wu integral" because it is based on the finding that for a fuel with a Lewis number of 1, the temporal changes in the normalized fuel mass fraction and normalized temperature in zero-dimensional homogeneous ignition are equivalent to the spatial changes in a one-dimensional laminar premixed flame. In other words, according to this embodiment, it is possible to provide a knocking prediction method and knocking prediction device that can further improve accuracy.

[0094] Furthermore, the knocking prediction method and knocking prediction device according to the present embodiment have high knocking prediction accuracy, and therefore bring great benefits to the design of combustion devices, power generation devices (internal combustion engines), vehicles, fuels, and the like.

[0095] For example, in the field of combustion equipment design, it is necessary to sacrifice some degree of heat energy recovery rate in order to avoid equipment damage due to knocking, and in the field of power generation equipment (internal combustion engine) and vehicle design, it is necessary to sacrifice some degree of output power in order to avoid equipment damage due to knocking.

[0096] In the design field of such combustion devices, power generating devices (internal combustion engines), vehicles, etc., improved knocking prediction accuracy makes it possible to reduce the thermal energy recovery rate and the sacrifice of output. In other words, the knocking prediction method and knocking prediction device according to the present embodiment make it possible to estimate the margin for knocking that should be considered in the design of combustion devices, power generating devices, vehicles, etc., with a narrower margin than ever before.

[0097] Therefore, by designing a combustion device, a power generating device (internal combustion engine), a vehicle, etc. based on the knocking occurrence conditions estimated by the knocking prediction method and knocking prediction device according to this embodiment, it is possible to realize a design that improves the performance of the combustion device, the power generating device (internal combustion engine), a vehicle, etc.

[0098] According to the knocking prediction method and knocking prediction device of this embodiment, it is possible to provide a method for designing a combustion device, a method for designing a power generation device, and a method for designing a vehicle that can improve the performance of a combustion device, a power generation device, a vehicle, etc.

[0099] In such a combustion device design method, power generating device design method, and vehicle design method, it is conceivable that fuels having a Lewis number greater than 1 are to be predicted (calculated). That is, as shown in the characteristic diagram of FIG. 4, fuels having a Lewis number less than 1 are unlikely to cause knocking, and therefore may be excluded from prediction (calculation) targets.

[0100] Furthermore, the knocking prediction method and knocking prediction device according to the present embodiment can accurately predict knocking for each type of fuel, which can contribute to the design of fuels that avoid knocking. Therefore, the knocking prediction method and knocking prediction device according to the present embodiment can provide a fuel setting method that can improve the performance of the fuel.

[0101] In this embodiment, the effectiveness of the present disclosure has been described for single-component fuels (pure substance fuels) such as "hydrogen" and "n-heptane" as examples of fuels to be evaluated for knocking prediction. However, the present disclosure can also be applied to other pure substance fuels such as methane and propane.

[0102] Furthermore, this embodiment predicts knocking of the fuel to be evaluated based on a one-dimensional steady-state analysis (simulation) of the fuel to be evaluated. Therefore, the present disclosure is not limited to pure-substance fuels, and can be applied to multi-substance fuels that are mixtures of multiple components, as long as one-dimensional steady-state analysis (simulation) is possible.

[0103] As is well known, in order to perform one-dimensional steady-state analysis (simulation) of fuels, it is necessary to identify the chemical species involved in the fuel combustion phenomenon and the elementary reactions based on those chemical species. For example, fuels for which chemical species and elementary reactions have been identified at present include surrogate fuels consisting of up to five components defined in the SIP (Surrogate Innovative Combustion Technology) surrogate chemical reaction mechanism. However, for multi-component fuels consisting of six or more components, the chemical species and elementary reactions have not yet been identified in the SIP surrogate chemical reaction mechanism.

[0104] However, it is fully expected that the chemical species and elementary reactions of multi-substance fuels consisting of six or more components will be identified in the future. In other words, it is fully expected that the chemical species and elementary reactions of multi-substance fuels consisting of six or more components will be identified and one-dimensional steady-state analysis (simulation) will become possible, and it is estimated that knocking prediction based on the present disclosure will be fully possible.

[0105] The present disclosure can be utilized in knocking prediction methods, knocking prediction devices, combustion device design methods, power generation device design methods, vehicle design methods, and fuel design methods.

[0106] A. Knocking prediction device 1. Storage unit 2. Calculation unit 3. Operation unit 4. Output unit

Claims

1. A knocking prediction method for predicting knocking that may occur in a premixed combustion field, the method predicting knocking based on the finding that in the premixed combustion field, the temporal changes in normalized fuel mass fraction and normalized temperature in zero-dimensional homogeneous ignition are equivalent to the spatial changes in a one-dimensional laminar premixed flame.

2. A knocking prediction method for predicting knocking that may occur in a premixed combustion field, comprising: a characteristic diagram showing the relationship between normalized temperature and normalized fuel mass fraction of a fuel; a fuel having a Lewis number greater than 1 is deemed to have the possibility of knocking; and a fuel having a Lewis number smaller than 1 is deemed not to have the possibility of knocking.

3. A knocking prediction method for predicting knocking that may occur in a premixed combustion field, comprising: a calculation step for sequentially calculating the combustion velocity when the inlet temperature is gradually increased by one-dimensional steady-state analysis of the premixed combustion field; a determination step for determining whether the combustion velocity has become impossible to obtain because a flame can no longer exist in the premixed combustion field; and a generation condition acquisition step for acquiring the inlet temperature corresponding to the last acquired combustion velocity as the generation condition for the knocking.

4. A knocking prediction method as described in claim 3, wherein in the calculation step, the size of the region including the unburnt mixture, the flame reaction zone and the burnt gas in the premixed combustion field is calculated by reducing it to the limit that can contain the flame.

5. A knocking prediction method according to claim 3 or 4, wherein the prediction target is a fuel having a Lewis number of greater than 1.

6. A knocking prediction device for predicting knocking that may occur in a premixed combustion field, comprising: a calculation means for sequentially calculating the combustion velocity when the inlet temperature is gradually increased by one-dimensional steady-state analysis of the premixed combustion field; a determination means for determining whether the combustion velocity has become impossible to obtain because a flame can no longer exist in the premixed combustion field; and an occurrence condition acquisition means for acquiring the inlet temperature corresponding to the last acquired combustion velocity as the occurrence condition of the knocking.

7. A knocking prediction device as set forth in claim 6, wherein said calculation means calculates the size of the region including the unburnt mixture, the flame reaction zone and the burnt gas in the premixed combustion field by reducing it to the limit that can be contained by the flame.

8. A knocking prediction device according to claim 6 or 7, wherein the calculation means performs calculations on fuels having a Lewis number of greater than 1.

9. A method for designing a combustion device, comprising the steps of: designing a combustion device based on knocking occurrence conditions estimated based on the knocking prediction method according to claim 3 or 4.

10. A method for designing a power generating device, comprising the steps of: designing a power generating device based on knocking occurrence conditions estimated based on the knocking prediction method according to claim 3 or 4.

11. A vehicle design method, comprising: designing a vehicle based on knocking occurrence conditions estimated based on the knocking prediction method according to claim 3 or 4.

12. A fuel design method for designing fuel based on knocking occurrence conditions estimated based on the knocking prediction method according to claim 3 or 4.

Images