Method and apparatus for operating a heat exchanger

Abstract

Method for operating a heat exchanger in the intake manifold of an internal combustion engine, - wherein the heat exchanger is subjected to a flow of medium during temperature control, which contains components that are deposited on and subsequently removed from a surface of the heat exchanger, resulting in a variable thickness (d S ) a layer of deposited components of the medium to be tempered on a surface of the heat exchanger and consequently the thermal resistance (M) of the heat exchanger changes during operation, so that control and / or regulation of the temperature of the medium to be tempered is affected, - where the medium to be tempered is a mixture of fresh air and exhaust gas, - where the control and / or regulation of the temperature of the medium to be tempered is model-based, - taking into account the change in thermal resistance (M) in model-based temperature control and / or temperature regulation, - where it is assumed that the thermal resistance (M) is composed of a constant component (M0) and a time-dependent component (M1(t)), where the time-dependent component (M1(t)) is further dependent on the variable thickness (d) S ) is dependent and the change in thickness (d S ) is attributed to the mass or quantity flow rate of the medium to be tempered, which flows through the heat exchanger, - wherein the deposition and removal of components of the medium to be tempered on a surface of the heat exchanger is described by a differential equation concerning the variable thickness (d S ) is described, - where a determination of parameters k appearing in the differential equation d , k R, which take into account the application to the surface of the heat exchanger and the removal from the surface of the heat exchanger during the operation of the heat exchanger.

Description

[0001] The present invention relates to a method and a device for operating a heat exchanger with the features of the claims.

[0002] For example, in vehicles with exhaust gas recirculation (EGR), a layer of soot forms on the gas-side cooling surfaces, particularly on the charge air cooler, due to particle deposition. From a thermal perspective, this deposited soot layer represents additional thermal resistance compared to a clean heat exchanger. In model-based temperature control algorithms, neglecting this effect leads to a decrease in control accuracy over the vehicle's operating lifetime. This can result in reduced power output and increased raw NOx emissions from an internal combustion engine. Therefore, over the vehicle's lifetime, the advantages of model-based temperature control, especially high control accuracy with minimal parameterization effort, are negated.

[0003] According to document US 8 028 569 B2, it is state of the art to determine and adapt the efficiency of an EGR cooler over its operating time so that the modeling of the temperature of the recirculated exhaust gas is more successful, using a reference cooler.

[0004] Furthermore, according to document DE 10 2018 116 983 A1, it is known to estimate the layer thickness of a cooling system's soot buildup. However, the thickness estimate is general and does not take essential influencing factors into account.

[0005] Modeling and verification of the effectiveness of an EGR cooler is also described in document US 8 725 386 B2, whereby adjustments to the model are made using certain factors to describe the cooler's changing behavior over time.

[0006] According to document US 2013 / 0312716A1, it is also known to monitor the effectiveness of an EGR cooler, among other things depending on the EGR mass flow, using a heat transfer coefficient, but a change in layer thickness is not taken into account.

[0007] According to document WO 2020 / 074 115 A1, in connection with determining the aging of a cooler, at least a layer of soot is quantified, whereby the disclosure is a purely diagnostic function and no adjustment of control parameters is provided. Rather, a rough numerical solution is described here.

[0008] Furthermore, according to document US 2014 / 0290631A1, it is known to determine the fouling of an EGR cooler by taking into account an estimated amount of solids in the exhaust gas and the thermal resistance caused by deposits, considering the thickness of the material from which the cooler is made, i.e., without fouling.

[0009] Document US 2014 / 0060503A1 further states that the efficiency of an EGR cooler can be determined from the characteristics of the EGR cooler with and without deposits.

[0010] The object of the present invention is to further improve the operation of a heat exchanger.

[0011] This problem is solved according to the invention by a method and a device having the features of the claims.

[0012] The inventive method for operating a heat exchanger enables the parameters for describing the heat exchanger's thermal resistance to be continuously adjusted, thus ensuring consistently high control accuracy over its service life and preserving the advantages of model-based temperature control. The programming effort is minimal, and the designed function can be executed in real time by a standard engine control unit. The present invention is not limited to application in an EGR cooler but also extends, in particular, to heat exchangers in the intake manifold of internal combustion engines with low-pressure EGR (LPG). A key difference between these cooler types is that, in the latter, in addition to soot buildup, a cleaning effect can also occur at very low LPG rates.Therefore, the sooting modeling for a heat exchanger in the intake path, e.g., of an intercooler, represents a generalization of the model of an EGR cooler, since in this case the EGR rate is (so to speak) always 100%.

[0013] Further embodiments of the present invention and a description of the advantages achieved can be found in the following exemplary embodiment and in the dependent patent claims.

[0014] With increasingly stringent emissions regulations and the growing competitiveness of alternative drive systems, targeted temperature control of individual vehicle components is becoming ever more important. According to the still unpublished DE 10 2020 116 218.9, a state-of-the-art method allows for demand-based temperature control of a working medium (intake air, exhaust gas) using heat exchangers. However, it is known that the ethylene glycol content in the cooling medium may change during operation of the heat exchanger. This can be addressed using a method described in DE 10 2016 124 652 B3. However, there are other effects that must be considered within the framework of model-based, demand-based temperature control.In combustion engines, this includes, for example, the fact that, due to exhaust gas recirculation, the heat exchanger(s) (especially the charge air cooler) are cooled by a medium or a mixture of several media, particularly fresh air and exhaust gas, which leads to increasing soot buildup. This, in turn, alters the characteristics of the heat exchanger. Neglecting this effect can result in a model stored in the engine control unit systematically deviating from reality. If, as proposed in DE 10 2020 116 218.9, a model-based control method is used, this can lead to reduced control accuracy compared to the target state. This can negate the advantages inherent in model-based control, particularly high control accuracy with minimal parameterization effort.

[0015] For this reason, the invention provides for the sooting effect to also be described in a model, in order to take it into account from the (motor) control or regulation side, i.e., within the context of influencing the temperature of a medium or medium mixture to be tempered or cooled. The basis for this is the dimensionless temperature change P of the heat exchanger. If the medium (to be cooled) (hereinafter also referred to as working fluid / medium) is brought into thermal contact with the (cooling) medium (hereinafter also referred to as cooling fluid / medium) in the heat exchanger, the dimensionless temperature change is defined by P=T1,us−T1,dsT1,us−T2,us, in which T 1,us and T 1,ds The temperatures of the working fluid upstream and downstream of the heat exchanger are respectively denoted. At T 2,usThis refers to the temperature of the cooling fluid upstream of the heat exchanger.

[0016] As is well known from the literature, P depends on the operating point. This means that the actual value of this quantity depends on the heat capacity fluxes Γ. j with j ∈ {1,2} of the two media, where index 1 denotes the medium to be cooled and index 2 the cooling medium. The heat capacity fluxes are given by the product of the specific heat capacity at constant pressure c. p,j and the associated mass flow ṁ j of the respective medium j. It therefore holds that Γj=cp,jm˙j.

[0017] Furthermore, the dimensionless temperature change also depends on the specific geometry of the heat exchanger. As a rule, this characteristic can be described by an effective heat transfer coefficient y, where the following applies: γ=11α1A1+M+1α1A2.

[0018] Herein, α describes j the heat transfer coefficient of the medium j and A j The area of ​​thermal transfer is defined by the coefficient of thermal transfer (C). The specific material properties of the heat exchanger are taken into account by the parameter M. In general, the two thermal surfaces A1 and A2 can differ, for example, if one side has a flat surface while the other side has fins or similar features.

[0019] It is possible here α j to write it as follows: αj=λjNujLj, wherein λ j represents the thermal conductivity of the medium j and L j a characteristic length of the system. With Nu jThe Nusselt number of the medium j is denoted. This can in turn be expressed in terms of the Reynolds and Prandtl numbers. As is known from the literature, for example according to A. Herzog, C. Pelka and F. Skorupa, Analytical Description of Thermal Control Circuits in Vehicles, in Energy and Thermal Management, Air Conditioning, Waste Heat Recovery 1st ETA Conference, December 1-2, 2016, Berlin, Germany, a formulation using the heat capacity flow rate is then possible. The following applies: αj=ujΓjmj in which u j depends only very weakly on temperature, so that this dependence can usually be neglected. A generalization to the case u j = u j (T j ) is easily possible. The explicit calculation of u j This is done by choosing a separation ansatz for the Nusselt number in which the Γ jThe -dependent component is separated. The proportionality constant between the heat transfer coefficient and Γjm is then through u j Given. For further details, see A. Herzog, C. Pelka and F. Skorupa, Analytical Description of Thermal Control Circuits in Vehicles, in Energy and Thermal Management, Air Conditioning, Waste Heat Recovery 1st ETA Conference, December 1-2, 2016, Berlin, Germany. j This is an adjustable parameter that can be determined depending on the heat exchanger.

[0020] Thus, y can be written as γ=aΓ1m1Γ2m2Γ1m1+bΓ2m2+MaΓ1m1Γ2m2, in which the following sizes were defined: a=A2u2 and b=aA1u1.

[0021] The parameterization of the model for the effective thermal conductivity y of the heat exchanger can therefore be done by measuring it.

[0022] However, if the heat exchanger is subjected to a flow of a medium / mixture of media (especially air and combustion gas) containing components that can accumulate on the (inner) wall of the heat exchanger, i.e., if solid and / or liquid components of the exhaust gas from an internal combustion engine (e.g., soot particles, tar, etc.) are present, then the value M is not a constant parameter over the component's life cycle. Rather, M changes over time. It is advantageous to take this change into account when using model-based temperature control.

[0023] For this purpose, according to the invention, the quantity M, i.e., the thermal resistance, is split into a constant component M0 and a time-dependent component M1(t), which is intended to account for the gradual fouling of the cooler. Therefore, the following applies: M(t)=M0+M1(t).

[0024] From the theory of heat transfer, it is known that M0 can be written as M0=dWλWAW, in which d W and A W Let w denote the thickness and thermal surface area of ​​the heat exchanger material, respectively. The thermal conductivity of the material is denoted by λ. W designated.

[0025] By complete analogy, an approach can be chosen for the layer s caused by the sooting. Maintaining the same notation, the following then follows: M1=dSλSAS.

[0026] According to the invention, a method is now provided which makes it possible to model the time-dependent part or the time-dependent thermal resistance M1 within the framework of an algorithm that can be implemented on an engine control unit.

[0027] The modeling of y, as specified in equation (2), is state of the art for M1(t) = 0. According to the invention, consideration is also given to the additional thermal resistance due to the soot layer or deposit resulting from exhaust gas recirculation through a heat exchanger, and a resulting engine control unit-compatible adaptation algorithm.

[0028] Assuming a homogeneous distribution of the soot buildup in the heat exchanger, it immediately becomes clear that λ S (thermal conductivity of the layer) and A S (thermal surface area of ​​the layer) is a constant. Consequently, the increase in heat transfer resistance depends on the layer thickness d. Stogether. The formation of the layer can be seen as an interplay of deposition and erosion. Using a molecular model for the deposition and erosion of particles on the air-side cooler wall, the layer thickness can be determined by a differential equation for d S can be described. Particular importance is attached to the mass flow rate of the working fluid to be cooled (fresh air or fresh air and exhaust gas). According to the invention, the development of the layer thickness d is described. S This mass flow is attributed to or dependent on this mass flow, which is advantageous because this mass flow represents a quantity calculated in the (engine) control unit and is therefore available for further processing.

[0029] The general case where the working fluid to be cooled, which flows through the heat exchanger, consists of a fresh air mass flow ṁ L (t) and an exhaust gas mass flow ṁ A(t) can be described by the following differential equation ddS(t)dt=kdm˙A(t)−kR[m˙A(t)+m˙L(t)]3dS(t).

[0030] The parameters k that appear here and are to be determined d and k R These calculations take into account the application and removal of solid and / or liquid components of the working fluid to and from the surface of the heat exchanger. As already mentioned, it is assumed that ṁ L and ṁ A This information is available either through sensors or through models implemented in the engine control unit, as is state-of-the-art. The dependence on the fresh air mass flow only exists in the case of a heat exchanger in the vehicle's intake manifold (intercooler). Of course, the method is also applicable to an EGR cooler. In such a case, ṁ L = 0.

[0031] The solution to the differential equation leads to dS(t)=kde−kR∫0t[m˙A(t1)+m˙L(t1)]3dt1∫0tm˙A(t2)ekR∫0t2[m˙A(t3)+m˙L(t3)]3dt3dt2.

[0032] Here, t1, t2 and t3 correspond to the integration variables and e to Euler's number.

[0033] If one defines the constant kD:=kdλSAS, The time-dependent thermal resistance of the soot layer can thus be written as M1(t)=M1[m˙A,m˙L]=kde−kR∫0t[m˙A(t1)+m˙L(t1)]3dt1∫0tm˙A(t2)ekR∫0t2[m˙A(t3)+m˙L(t3)]3dt3dt2.

[0034] Thus, the effective heat transfer coefficient is functionally given by γ=γkD,kR(Γ1,Γ2,m˙L,m˙A).

[0035] According to the invention, it is advantageous to adjust the parameters k D and k R during the operation of the internal combustion engine or the underlying vehicle. That is, the parameters k, which are assumed to be constant. D and k RThese values ​​can now be evaluated for a specific heat exchanger and implemented in the control unit based on the associated dimensionless temperature change. However, this process is very time-consuming, as sooting must be induced for each model of the heat exchanger in question, which entails lengthy measurement series. For this reason, an algorithm is preferable that allows the engine control unit to determine the additional thermal resistance caused during operation. The basis for this is again the dimensionless temperature change, which, according to the explanations above, is... P[Γ1,Γ2,γ(Γ1,Γ2,M(t))] can be formulated.

[0036] Assuming that a local inverse of this function with respect to y exists, which is the case for common heat exchangers, the following also applies: γDTA=γDTA(Γ1,Γ2,P)

[0037] Here, P is defined according to (1) when the temperatures T are known. 1,us , T 1,ds and T 2,us The value is available as a numerical value based on sensors and models. The index DTA indicates that the heat transfer in this context was determined based on the dimensionless temperature change. Conversely, γ also exists within the framework of heat conduction theory according to equation (5).

[0038] During the parameterization process of a new vehicle, y is measured according to equation (2) with M1(t = 0) = 0 as a standard procedure and parameterized in the engine control unit. According to this invention, increasing soot buildup is then addressed as follows during driving operation.

[0039] A comparison is made between the sensorily accessible value γ DTA and the value y present in the (engine) control unit for the respective operating point. If the sensor-detected value γ now deviates DTAsystematically from the model value y, an on-board adjustment of the parameters k should be performed. D and k R This is done by the control unit. Since the model for the dimensionless temperature change is a steady-state description of the heat exchanger, the adaptation algorithm to be described is only used if the engine control unit has established a fixed operating point for the heat exchanger. Fulfillment of this condition is detected by falling below the rate of change of P. If the sampling time in the control unit is denoted by τ, then the algorithm is triggered if the condition is met. ∏≥|P(t+τ)−P(t)τ| is satisfied, where Π represents a limit value indicating steady-state behavior with respect to the thermal state variables.

[0040] If the above condition is met, then the adjustment of the parameters k D and k Rwithin N time steps, the solution of a nonlinear least squares problem is explicitly defined by kD,kRmin{∑j=1N[γDTA(j)−γkD,kR(j)(Γ1(j),Γ2(j),m˙L(j),m˙A(j))]2} This has been formulated. Once this is solved, a way has been found to account for the sooting effect in the model. This involves adapting the parameters k. D and k R during operation, which conversely also results in a consistently high level of control quality throughout the entire operating time of the vehicle.

[0041] Equation (8) can be solved in different ways. One possibility is to use the values ​​already used. kD(0) and kR(0) a linearization of γkD,kR(j)(Γ1(j),Γ2(j),m˙L(j),m˙A(j)) to use.

[0042] In summary, the adaptation algorithm then proceeds as follows:

[0043] 1. If condition (7) is met, the following sum is formed over a given period t = Nτ with N sampling steps: σ=∑j=1N|γDTA(j)−γkD,kR(j)(Γ1(j),Γ2(j),m˙L(j),m˙A(j))|.

[0044] This involves γDTA(j) according to equation (6) and γkD,kR(j) calculated according to equation (5). In particular, the following applies: γkD,kR(j)(Γ1(j),Γ2(j),m˙L(j),m˙A(j))=aΓM1m1(j)ΓM2m2(j)ΓM1m1(j)+bΓM2m2(j)+M(j)aΓM1m1(j)ΓM2m2(j) in which the material properties also depend on the respective time step according to equation (3).

[0045] The parameters a, b, m1, m2 and M0 are determined during the parameterization of the model-based control approach. The time-dependent function M1(t), which describes the sooting, represents a central element of the invention (4), enabling the sooting to be adjusted over the vehicle's lifetime.

[0046] The corresponding values ​​for γDTA(j) and γkD,kR(j) In equation (9), the parameters are stored by the engine control unit to aid in the adaptation of the parameters k. D and k R to be able to access it.

[0047] 2. If σ exceeds a critical value Σ, the adaptation for the parameters k D and k R started. Condition (8) initially leads to 0=∑j=1N[γDTA(j)−γkD,kR(j)]Γ1m1(tj)Γ2m2(tj)[Γ1m1(tj)+bΓ 2m2(tj)+aΓ1m1(tj)Γ2m2(tj)(M0+M1(kD,kR))]2∂M1(kD,kR)∂kL with L ∈ {D, R}.

[0048] Now we would like to add to this expression the existing values kD(0) and kR(0) linearize.

[0049] 3. In the next step, we define the function fL(kD,kR)=∑j=1N[γDTA(j)−γkD,kR(j)]Γ1m1(tj)Γ2m2(tj)[Γ1m1(tj )+bΓ2m2(tj)+aΓ1m1(tj)Γ2m2(tj)(M0+M1(kD,kR))]2×∂M1(kD,kR)∂kL and define ξ:=kD−kD(0), as well as ζ:=kR−kR(0).

[0050] This results from equation (10) 0=fL(kD(0)+ξ,kR(0)+ζ).

[0051] A linear development then yields with fDDξ+fDRζ=−fD(0) and fRDξ+fRRζ=−fR(0) a system of linear equations, using the following notations: fL(0)=fL(kD(0),kR(0)) and fXY:∂fX(kD,kR)∂kY|kY=kY(0).

[0052] The solution to the system of equations is given by ξ=fR(0)fDR−fD(0)fRRfRRfDD−fDRfRD and ζ=fD(0)fRD−fR(0)fDDfRRfDD−fDRfRD Given. Together with equations (11) and (12), this yields updated values ​​for the parameters k. D and k R .

Claims

[1] Method for operating a heat exchanger in the intake section of an internal combustion engine, - wherein the heat exchanger is subjected to a flow of medium during temperature control, which contains components that are deposited on and subsequently removed from a surface of the heat exchanger, resulting in a variable thickness (d S ) a layer of deposited components of the medium to be tempered on a surface of the heat exchanger and consequently the thermal resistance (M) of the heat exchanger changes during operation, so that control and / or regulation of the temperature of the medium to be tempered is affected, - where the medium to be tempered is a mixture of fresh air and exhaust gas, - where the control and / or regulation of the temperature of the medium to be tempered is model-based, - taking into account the change in thermal resistance (M) in model-based temperature control and / or temperature regulation, - where it is assumed that the thermal resistance (M) is composed of a constant component (M0) and a time-dependent component (M1(t)), where the time-dependent component (M1(t)) is further dependent on the variable thickness (d) S ) is dependent and the change in thickness (d S ) is attributed to the mass or quantity flow rate of the medium to be tempered, which flows through the heat exchanger, - wherein the deposition and removal of components of the medium to be tempered on a surface of the heat exchanger is described by a differential equation concerning the variable thickness (d S ) is described, - where a determination of parameters k appearing in the differential equation d , k R, which take into account the application to the surface of the heat exchanger and the removal from the surface of the heat exchanger during the operation of the heat exchanger. [2] Method for operating a heat exchanger which is a cooler for the exhaust gas of an internal combustion engine, - wherein the heat exchanger is subjected to a flow of medium during temperature control, which contains components that are deposited on and subsequently removed from a surface of the heat exchanger, resulting in a variable thickness (d S ) a layer of deposited components of the medium to be tempered on a surface of the heat exchanger and consequently the thermal resistance (M) of the heat exchanger changes during operation, so that control and / or regulation of the temperature of the medium to be tempered is affected, - where the medium to be tempered is exhaust gas, - where the control and / or regulation of the temperature of the medium to be tempered is model-based, - taking into account the change in thermal resistance (M) in model-based temperature control and / or temperature regulation, - where it is assumed that the thermal resistance (M) is composed of a constant component (M0) and a time-dependent component (M1(t)), where the time-dependent component (M1(t)) is further dependent on the variable thickness (d) S ) is dependent and the change in thickness (d S ) is attributed to the mass or quantity flow rate of the medium to be tempered, which flows through the heat exchanger, - wherein the deposition and removal of components of the medium to be tempered on a surface of the heat exchanger is described by a differential equation concerning the variable thickness (d S ) is described, - where a determination of parameters k appearing in the differential equation d , k R , which take into account the application to the surface of the heat exchanger and the removal from the surface of the heat exchanger during the operation of the heat exchanger. [3] Method according to claim 1 or 2, wherein the constant component of the thermal resistance (M0) is composed of M0=dWλWAW, results, where d W , A W and λ W w is the thickness, thermal surface area and thermal conductivity of the heat exchanger material, where the time-dependent component (M1(t)) is derived from M1=dSλSAS. this results in, where d s , A s and λ s These are the thickness, the thermal surface area, and the thermal conductivity of the layer of deposited components of the medium to be tempered. [4] Method according to claim 1, wherein the heat exchanger is an air cooler of an internal combustion engine and the medium to be tempered consists of a fresh air mass flow ṁ L (t) and an exhaust gas mass flow ṁ A (t) is composed and the variable thickness (d) S ) by differential equation ddS(t)dt=kdm˙A(t)−kR[m˙A(t)+m˙L(t)]3dS(t) is described, where the parameters k d and k R describe the deposition and removal of components of the medium to be tempered on a surface of the heat exchanger. [5] Method according to claim 2, wherein the medium to be tempered is only an exhaust gas mass flow ṁ A (t) includes a fresh air mass flow ṁ L (t) is set to zero and the variable thickness (d) S ) by differential equation ddS(t)dt=kdm˙A(t)−kR[m˙A(t)+m˙L(t)]3dS(t) is described, where the parameters k dand k R describe the deposition and removal of components of the medium to be tempered on a surface of the heat exchanger. [6] Device which is configured to carry out one of the methods according to claims 1 to 5. [7] Vehicle comprising a device according to claim 6.