"Efficiency for the Field," Feature Article, April 2007

efficiency for the field

A manufacturer optimizes cooling in the confines of a tractor engine.

by Panos Tamamidis

the mechanization of agriculture stands as one of the leading achievements of engineering in the 20th century. As the internal combustion engine took over more work formerly done by sheer muscle, the world's harvests reached unprecedented abundance.

The mechanized workhorse of agriculture is the tractor. It carries out the most basic of jobs: It pulls the plow. What's more, it moves and powers numerous other implements of contemporary farming.

It is a peculiar characteristic of tractors, because of the way they are used and the way they must be designed, that keeping their various cooling systems efficient can be a difficult challenge. The engine may not be so different from one powering a truck down the Interstate, but the environment and the workload are very different.

A large, turbocharged diesel tractor can have five individual cooling systems under the hood—one for the engine, another for fuel, a third for the compressed air on its way to the cylinder, another for transmission oil, and a fifth for cabin air conditioning. The length, height, and width of a tractor's engine compartment are smaller than those of a truck with a comparable engine because the driver of a tractor needs to see the ground closer to the wheels than does the driver of an 18-wheeler.

What's more, the tractor, with an estimated plowing speed of 6 kilometers an hour, gets no advantage from wind speed. At least, it gets nothing like the airflow to a 200-hp diesel doing 65 miles an hour on the highway.

Instead, the tractor has to rely almost entirely on a fan to carry off heat rejected by its cooling systems. There is room for only a single fan, which like all the other systems in the tractor, takes its power from the engine.

As a manufacturer of construction and agricultural equipment, CNH Case New Holland has plenty of experience dealing with the issues of heat transfer in tractor engines.

CNH builds a wide range of equipment, which is sold in about 160 countries around the world. The trend in today's commercial farming is toward large engines that present greater cooling challenges and frequently also take up more space, which often means less room for the cooling package. New Tier III and IV emissions regulations create additional cooling challenges.

Increasing the efficiency of the cooling package is a major issue, not only because of size, but also because the fan draws power that would otherwise be available to drive the tractor and implements, and it affects fuel economy.

The fan must deliver enough air to each module to meet its cooling requirements while drawing the least possible amount of power from the engine. The design process often requires finding the best placement of the different modules inside the engine compartment to provide optimum cooling at the lowest cost.

Modeling airflow

At one time, the company had to build and test a prototype to test each arrangement. Now, engineers at CNH Case New Holland have developed a method that involves the use of computer simulation to model airflow through the cooling system, making it possible to predict the performance of proposed designs without the expense of building a prototype for each one. In a recent product launch in the company's series of Magnum tractors, these methods made it possible to evaluate enough potential designs to optimize the design so that fan power is reduced significantly. They also reduced the costs of building and testing prototypes.

In the past, CNH engineers performed engineering calculations that made it possible to estimate the total airflow provided by the fan and the airflow required by each module. The calculations didn't take into account the geometry of the underhood compartment and so didn't demonstrate how the airflow was distributed. This uncertainty made it necessary to build and test prototypes to find a design that met the cooling and space requirements. The number of design iterations was not nearly enough to optimize a design.

The new method developed by CNH engineers uses computational fluid dynamics to simulate the flow of air from the fan, through the engine compartment, and through the heat exchangers used in each of the modules. CNH used software from Fluent Inc. in Lebanon, N.H., because it has demonstrated the ability to accurately simulate extremely complicated cooling packages while keeping computational requirements to reasonable levels.

Engineering data included the computer-aided design geometry from the manufacturer of the engine, Consolidated Diesel, a joint venture of Case New Holland's parent, CNH Global, and Cummins Engine Co. Information was also collected for heat exchangers, fan, and other components, as well as the engine compartment sheet metal.

In the past, calculations accounted for the fan based on its performance specifications as measured in a laboratory. The fan behaves differently under a hood. Using multiple frames of reference in the model made it possible to take into account the geometry of the underhood compartment and study the impact of different blade designs.

The frame of the fan blades and hub is a rotating one; that of the engine compartment is stationary. The solution proceeded with a steady transfer of information across a predefined interface between the two frames.

Engineers used CFD to simulate flow of air around the tractor, from the fan, through the engine compartment, and through the heat exchangers. Simulation also predicted the transfer of heat through the engine compartment.

The flow of coolant through the heat exchanger and various engine systems was not addressed by the simulation because it would add substantially to the already large computational requirements. Instead, the performance curves provided by the manufacturers of the heat exchangers were used to determine the performance of each cooling module based on the air-side pressure drop predicted by the CFD simulation. In order to reduce computational time, engineers modeled the heat exchangers as a porous region instead of duplicating every detail of their geometry.

The engineers also considered temperature requirements for fluids. For example, the engine coolant had to be maintained below a certain number in order to avoid creating steam, which can damage the water pump and derate the engine.

The engine was modeled as a black box based on data supplied by the engine manufacturer as inputs to the heat exchangers. The simulation also modeled the flow of air through the grille and into and out of the engine compartment. The resulting model had between 5 million and 10 million elements and took between 24 and 48 hours to solve on a server with more than four processors.

The initial design showed oil temperatures too high. Engineers changed the cooling package configuration and evaluated different oil coolers. The best configuration they found put the charge-air cooler for the intake manifold and the radiator in parallel. This approach reduced the oil temperature to acceptable levels and made it possible to reduce the size of the radiator and charge-air cooler, which saved space inside the engine.

Engineers spent a lot of time optimizing the design of the fan. They simulated several different blade designs seeking a design that would reduce the power draw. They adjusted the position of the different modules. They added louvers in the engine compartment to let air move freely and reduce the amount of work performed by the fan. The engineers were able to complete each new design iteration quickly, making it possible to evaluate far more design alternatives than would have been possible if they had used the build and test approach.

The result was that CNH engineers were able to significantly reduce the amount of power drawn by the fan compared to the previous version while meeting all cooling module temperature and packaging requirements. Coolant temperatures were also lower than in the previous design.

The next step was building a prototype of the design that was defined through simulation. Coolant temperatures measured in physical testing of this prototype correlated very well with those predicted by the simulation. These results were considered excellent because they approximated the variability of physical testing. The use of simulation in the design process made it possible to bring the product to market substantially faster than the previous generation of products.

Panos Tamamidis is CFD manager for CNH Case New Holland in Burr Ridge, Ill.