One of the most difficult practical problems to overcome when testing agricultural machines (tractors) at the dyno is measuring the fuel consumption and assessing the engine’s global efficiency. Indeed, diesel engines have two fuel-lines: a delivery line (from the tank to the cylinders), and a return line, to convey back in the tank the fuel not injected as a consequence of the accelerator position and pump settings. In principle, to measure the consumption, known the delivery flowrate (proportional to the engine speed), a single flowmeter could be used in the return line only (having a variable flow); unluckily, the real suction flowrate is generally unknown, varying with the model and vehicle age (i.e., with the pump components’ wear), so two flowmeters are always needed. But there is another metrological problem to consider: a flowmeter inserted in series in a fuel-supply line could significantly alter the flowrate. Finally, another further problem can arise, when the experimenters are asked to characterize machines not owned by them: modifications to the vehicle under test should be avoided or, at least, limited, e.g., not to void the guarantee. Unfortunately, solutions like the clamp-on ultrasonic flowmeters (placed outside the pipes) are not so suitable for fluids with few particles in suspension, e.g., the fuel processed by the pump. Therefore, it was decided to act upstream of the entire supply line, providing an external tank placed on a precision scale near the tractor, to be used instead of the vehicle’s tank. The only, absolutely reversible, needed operation is to connect to the fuel pump two flexible pipes drawing from the external tank. The device gives the net fuel consumption by means of a data logger that acquires the tank mass at periodic intervals. It has been successfully tested on a New Holland 4020V tractor at the dyno, allowing to calculate the instant/hourly/specific consumptions and the engine’s global efficiency at six engine speeds (from 800 to 2200 rpm) and full throttle.
