The exhaust heat recovery system (EHRS) in an automobile captures the thermal energy from exhaust and transfers it to the engine coolant. As the car warms up, engine friction is reduced and transmission efficiency increases. Certain vehicles with idle speed bump, transmission oil heaters, and cold-temperature fuel enrichment will have enhanced impact of a faster warm-up. Therefore, the impact of the EHRS is very specific to a vehicle and must be evaluated separately. A method was developed to evaluate the effect of an EHRS on a 1.2-L, naturally aspirated, gasoline engine passenger car typical in India.

Analysis of an automobile as a complete system enables engineers to understand the effects of various technologies that can be employed for fuel economy and emission improvement. The benefits gained over standard duty cycles and for real-world driving conditions can be quantified. A study investigated the effect of an EHRS on a Modified Indian Drive Cycle (MIDC).

A forward-calculating model of the passenger car created in Simulink® (MathWorks, Natick, MA) was used to calculate the engine loading and heat rejection as well as the exhaust energy being generated. A calibrated model was used to simulate the MIDC, which was closely integrated with a transient, underhood thermal model to evaluate the effect of the EHRS on warm-up and engine friction reduction.

Vehicle level modeling allows manufacturers to optimize and iterate quickly. A complete correlated vehicle energy-balance model was developed in PowertrainLive® using a forward-looking approach for analysis with the thermal simulation software. The forward-looking model includes a driver model that adjusts the throttle to affect the vehicle speed. The vehicle speed is a function of the torque generated by the engine. The driver model is a combination of a PI controller, a look-ahead feature, and feed-forward controller. The forward-looking model is more appropriate for scenarios where controls and fast-transient phenomenon are important to study, such as the transmission shift map or a torque converter lock-up strategy.

Figure 1: Schematic of the vehicle model integrated with the underhood transient thermal model.

This model was closely coupled with a transient underhood thermal model of the cooling system and the engine lubrication system (Figure 1) to compute the warm-up performance and the difference in the warm-up times caused by the EHRS. The primary purpose of the vehicle thermal model was to capture the engine, coolant, and oil system's inertia and the change in warm-up from the heat load of the WHR system. A baseline engine thermal model was developed and correlated with the vehicle warm-up data. The appropriate corrections for change in engine friction as a result of the higher temperatures were also captured.

The coolant and engine oil flow systems were modeled using standard representations of the passages, bends, and other restrictions. The coolant system consisted of a thermostat valve opening system, cabin heater interaction system, and pump that controlled the coolant flow in-system. The engine thermal system consisted of engine thermal inertia and combustion heat interaction among various engine parts. The engine mass was lumped into a single mass to keep the model simple and fast in computations.

The engine oil system consisted of the heat interaction among various engine parts and the engine oil. The interactions between the fluids and the engine structure were modeled using thermal bridges that implement equations for convection and conduction heat transfer based on the coefficients provided. The complete 1D engine thermal model, created in FloMASTER (Mentor Graphics Corp., Wilsonville, OR) is shown in Figure 2.

Figure 2: The vehicle thermal model created in FloMASTER.

This engine thermal model was imposed with various boundary conditions such as vehicle speed, engine torque, and engine RPM vs. time from a larger vehicle simulation model. Because this is a transient thermal model, these boundary conditions are a trace of time for a particular drive cycle. The engine speed and brake torque were used by the engine thermal model to look up the instantaneous combustion heat, frictional heat, coolant, and oil flow rates and to then calculate the temperatures in various subsystems. The key outputs of the model were the transient temperatures of the coolant and engine oil, in addition to the thermal performance of individual components such as the radiator, engine oil cooler, etc.

The combustion heat rejected to the engine structure was calculated in PowertrainLive, which has a built-in algorithm that estimates engine heat rejection, exhaust mass flow, and temperatures. The engine friction was also available as a calculated parameter from PowertrainLive. The engine friction was estimated using a proprietary algorithm based on the original Heywood correlation, but improved to reflect and correlate with the latest generation of engines. This improved empirical correlation is based on engine speed, torque, oil viscosity, and oil temperature. As was the case with the combustion heat, the frictional heat (FMEP) was available as a function of engine RPM and BMEP, and adjusted based on the viscosity of the oil at the instantaneous temperature.

The EHRS was modeled as a heat load in the cooling circuit. The effectiveness of the EHRS was a function of coolant and exhaust mass flows and was observed to be ~85% cycle averaged, and was turned on until the engine coolant temperature reached 90 °C. Above that, the thermostat opened, indicating that there was excess heat in the system that needed to be rejected. At this point, the EHRS was counter-productive and thus bypassed. Calculations show that the thermal energy captured by the EHRS systems is ~28% of exhaust energy in the MIDC drive cycle. With the EHRS turned on, the engine oil warm-up sped up significantly. Analysis showed the engine warm-up time reduced from 850 to 550 seconds.

The EHRS improved the rate at which engine oil and coolant temperature warms up. This helped in reducing engine friction. This vehicle did not have any of the other calibration parameters implemented such as idle bump or a transmission oil heater. Given the absence of other cold-temperature calibration parameters, the primary fuel economy improvement was due to reduced cold-start engine friction. Using this method, it was observed that adding an EHRS improves engine oil and coolant warm-up, which results in reduced friction during cold start.

This work was done by Sudhi Uppuluri, Ajay M. Naiknaware, and Hemant Khalane of Computational Sciences Experts Group, Ann Arbor, MI. For more information, visit here.