This article also appears in
Subscribe now »

Parallel layout of the cooling loops in the prototype vehicle. The low temperature loop is for A/C condenser, turbo intercooler and other heat exchangers ("other loads") in the system. (To see more images, click on the small arrow at the upper right corner of this image.) 

Fiat demonstrates improved fuel efficiency with low-temperature liquid cooling loop

The concept of a second water-cooling loop—adding a low-temperature radiator (LTR) and water cooling to replace air-cooling of the A/C condenser, turbo air intercooler (charge air cooler) and other front-end heat exchangers—has been demonstrated by Fiat Research Center (CRF) engineers. The goal is to improve system thermal efficiency, which can translate into cost-effective improvements in fuel efficiency, with corresponding reduction in greenhouse gas emissions.

Originally indicated for A/C condenser cooling by modeling at several automakers, the second cooling loop as developed by the CRF has the LTR mounted ahead of the high-temperature radiator used for engine coolant. All individual heat exchangers on the LTR circuit are water-cooled and mounted as close as possible to the system components.

CRF estimates a 5% improvement in fuel economy during A/C operation on the New European Drive Cycle (NEDC), explained Roberto Monforte, R&D engineer in the thermal management area at CRF. He spoke with Automotive Engineering at a recent meeting of the SAE Interior Climate Control Committee.

Amesim model to real world

The project, for Fiat Chrysler Automobiles (FCA), began with a modeling exercise and expanded it to real-world level. CRF was able to show very close agreement between its model, which used LMS Amesim (a mechatronics simulation suite from Siemens) and actual measurements. They were taken during prototype development of FCA's Grande Punto subcompact car powered by a 1.3-L Multijet diesel rated at 95 hp (71 kW) and meeting Euro5 emission standards. The results should abet FCA Group efforts to engineer the dual-loop system into a wide range of cars with differing cooling requirements and number of air-cooled heat exchangers in their present configurations.

CRF has prototypes of other vehicles including three cars (an off-road model and a heavy-duty) but showed only the data for the Grande Punto.

Outwardly, the dual-loop system seems overly complex, and its use of an additional layer of heat exchange would appear to reduce efficiency. However, water-based liquid is a more effective heat exchange fluid than air and other forms of liquid cooling used on some cars for turbo air and engine oil. The present primary alternative—airflow cooling several different fluids, including engine and transmission oils, A/C refrigerant, power steering oil and turbo air—dictates that airflow from the fan must be able to handle the worst-case requirement. So even if just one front-end heat exchanger needs it, fan speed would be excessively high to handle that single demand.

With just one LTR in front, the temperature of the water-based fluid in that radiator (55-60⁰C/131-140⁰F) becomes an average of the entire low-temperature system. The fan power draw, using an infinitely-variable pulse-width-modulated fan, reflects relatively-consistent requirements for the water-based liquids of the engine coolant radiator (90⁰C/194⁰F) and the LTR . This results in a smooth power draw that is more efficient, and may also provide better engine-off cabin heating on vehicles equipped with engine top-start. Water pump flow is electronically controlled.

The engine cooling system lends itself to accurate modeling, as heat load is a function of engine rpm and BMEP (brake mean effective pressure). So does turbocharger air, as mass-flow rate and temperature are determined from a table that includes those two heat load entries.

Refrigerant charge reduction possible

The LTR might also seem to raise packaging issues. But only the LTR is located at the front, making the front-end cooling module simpler to install. Further, specific system heat exchangers may be packaged close to the system they're cooling, reducing length of the lines for the different oils, turbocharger air intercooler and A/C refrigerant. This lowers risk of physical damage and permits a smaller refrigerant charge.

While the CRF installation was done with R-134a refrigerant, a production version might be equipped with R-1234yf, which is so costly that even a 50-g (1.7-oz) reduction in charge would be worthwhile. Water cooling in conjunction with condenser sub-cooling indicated that an internal heat exchanger was not needed on the Grande Punto with R-134a; it would only be used in a system where considerably more sub-cooling is needed.

The Grande Punto required less than a more complex vehicle, as the only LTR loop heat exchangers are for the A/C condenser and turbocharger air intercooler, although evaporator performance also was part of the model. CRF engineers attempted to keep each of the heat exchanger models simple by using the sizing and performance data of the parts supplier.

Three condensers with sub-cool sections, from different suppliers, were evaluated. There was no specific size/configuration/performance data for the turbo intercooler or cabin heater arrangements, so the model took the NTU approach (Number of Transfer Units, a way to estimate heat transfer, particularly for counter-flow heat exchangers, when needed data from specific heat exchangers is not available).

Modeling results close to measured

Both the Amesim model and the Grande Punto were run on an NEDC, which begins with a hot start, and is run at 28⁰C. (82⁰F), 50% relative humidity. The A/C is on and set for 22⁰C/72⁰F; the system is in outside air and blower speed automatically controlled.

The Amesim model for the LTR and the actual measurement were, on average, within 2⁰C . For the water-cooled intercooler, the model was an average of about 3⁰C high on the NEDC's EUDC (extra-urban section, high-speed operation) segment. The model showed greater heat rejection than measured. And although the dynamic difference in that segment will be investigated, all the temperature levels were acceptable.

The loop for the condenser showed very close agreement between the Amesim model and the measurements, both for coolant temperature and R-134a condenser outlet pressure.

CRF also modeled and tested A/C cool-down at high ambient temperatures, beginning with a hot soak at 43⁰C (109⁰F), 30% relative humidity, solar load of 900 W/m². The Amesim cabin temperature and A/C refrigerant pressure tracked almost exactly with actual measurements, with only minor differences at engine idle.

Continue reading »