Fig. 1 shows the layout of a continuous air flow test bench [11]. It is composed by the following devices:
•
A screw compressor with a maximum mass flow capacity of 0.2 kg s−1, at a maximum discharging pressure of 3.5 bar (gauge), which provides the mass flow to the turbine. Mass flow rate is controlled by changing the screw compressor speed or the opening of an electronic discharge valve (placed downstream the screw compressor). This valve is used when a lower mass flow than the minimum supplied by the screw compressor is required being discharged to the atmosphere the extra flow.
•
Mass flow is heated in parallel using five tube-type electrical heaters, mass flow through each of the heaters can be regulated by means of a valve placed at their inlet ports. This system can reach up to 720 K at the maximum mass flow rate, this hot flow is collected later in a plenum and conducted to the turbine inlet.
•
After passing through the turbine, the air is cooled by means of a heat exchanger in order to allow mass flow measurement using high accuracy hot film flow meter. All flow meters in the installation have been previously calibrated.
•
Turbo-compressor sucks air from the atmosphere, air passes first through a filter before being measured. Downstream the compressor, an electronically driven back-pressure valve has been installed in order to emulate what would be engine intake valves. Hereinafter compressor refers only to the turbo-compressor.
•
An independent lubrication system is installed to control oil flow rate and its inlet pressure (by means of an oil pump and a controlled pressure valve). Temperature can be also controlled and modified as desired by using an electrical heater and a cooler. Oil mass flow rate is measured by means of a Coriolis flow meter, meanwhile inlet and outlet temperatures are measured using platinum resistance temperature detectors. Oil samples are taken periodically in order to characterize its properties (viscosity, density and specific heat capacity variations with temperature).
•
Temperature and pressure sensors are installed on the inlet and the outlet pipes of the compressor and the turbine according to SAE (Society of Automotive Engineers) J1723 [25] and SAE J1826 [26] standards. In this way the obtained results would be applied very quickly on any turbocharger previously measured following these standards that is usually performed in industry. This fact will be very interesting for both engine and turbocharger manufacturers. The methodology employed and the obtained results in this work could be used in other turbochargers previously tested following these standards which give an interesting and non-negligible tool for both researchers and industry.
Schematic test bench and location of main sensors.
Fig. 1.
Schematic test bench and location of main sensors.
Figure options
Table 1 shows main information about measurement range and uncertainty of sensors used in the test bench. Tests performed on this flow rig have been divided into two main groups, named as:
1.
Almost-adiabatic tests [9]. Whose main objective is to decouple mechanical losses characterization from heat transfer problem in the turbocharger under study [27]. In this way, heat transfer is minimized by setting turbine inlet, compressor outlet and lubricating oil inlet temperatures at the same level. Doing so lubricating oil enthalpy drop corresponds to the direct measurement of turbocharger mechanical losses that will be characterized and modelled using an empirical [28] or a physical model [11].
2.
Hot tests [10]. Main objective of this kind of tests is to characterize convective heat fluxes inside the turbocharger. Besides, these tests can be divided into two main groups:
(a)
Externally insulated tests. In these tests the whole turbocharger is externally insulated avoiding heat fluxes to the environment, appearing only internal heat fluxes.
(b)
Exposed tests. These tests are the usually performed by manufacturers in order to obtain turbocharger maps. Main difference respect to previous tests comes from the fact that heat fluxes to the environment are allowed. Environment conditions (temperature) and air flow through the bench have been measured in order to estimate accurately these heat fluxes.
Table 1.
Characteristics of sensors employed in the test bench.
Variable Sensor Range/error
Gas pressure Piezoresistive [0–5] ± 0.025 bar
Air pressure Piezoresistive [0–2] ± 0.025 bar
Gas and air temp. K-type therm. −200–+1200 ± 2.2 °C
Gas and air flow Hot wire [0–720] ± 0.72 kgh−1
Oil pressure Piezoresistive [0–6] ± 0.025 bar
Oil temperature RTD [−200–+650] ± 0.15 °C
Oil flow Coriolis [0–100]kgs−1 ± 0.1%
Table options
In this work, externally insulated tests have been performed in order to obtain, internal convective coefficients and correlations. These internal heat transfer fluxes have been obtained according to the thermal model explained in Section 3.
In order to extend validity of this work, three different turbochargers have been studied, whose main characteristics are shown in Table 2. Performance maps from those turbochargers are observed in Fig. 2. These turbochargers are typical used in small-medium Diesel engines, so the obtained results could be applied to similar turbochargers, i.e. similar sizes and compression/expansion ratios. The knowledge of heat transfer in this engines could lead to an improvement of their performance by reducing energy losses.
Table 2.
Main characteristics of the employed turbochargers.
Parameter First turbocharger Second turbocharger Third turbocharger
Turbine wheel diameter [mm] 41 38 36.5
Compressor wheel diameter [mm] 49 46 40
VGT yes, vanes yes, vanes no
Water cooled yes No yes
Type of journal bearing fixed floating ring floating ring
Engine power [kW] 129 96 75
Engine type diesel diesel petrol
Displacement [l] 2.0 1.6 1.2
Table options
Measured points for convective characterization. Left side compressor maps and ...
Fig. 2.
Measured points for convective characterization. Left side compressor maps and right side turbine maps. Legends refer to VGT opening in %
Figure options
2.1. Uncertainty analysis
The uncertainty of a measurement is a parameter that characterises the dispersion of the values that could reasonably be attributed to the action of measuring. As it has been proposed in Ref. [29] the uncertainty estimation can be evaluated by using a statistical analysis of series of observations and by other means, such as manufacturers data. In the case of the measurements of current work both types of evaluations have been performed. In one hand, the standard deviation due to the repetitiveness of the measurement is calculated using Equation (1), where n is the number of measurements, View the MathML source is the arithmetical mean of these measurements and xi the measurement. On the other hand the standard deviation due to the inaccuracy of each sensor can be computed using manufacturer data on the probability distribution of the error or assuming uniform rectangular distribution of probability if only the bounds are given [29]. In this last case the standard deviation is calculated using Equation (2) from uniform rectangular distribution of probability, where a− and a+ are the lower and the upper limits of the sensor inaccuracy.
equation(1)
View the MathML source
Turn MathJax on
equation(2)
View the MathML source
Turn MathJax on
Finally the standard deviation, representing combined uncertainty is calculated using Equation (3), taking into account both of the previous effects. Furthermore, it is also used for the computation of the uncertainty of derived variables. Using these expression all the measured or computed variables of this paper are provided with uncertainty limits given in terms of standard deviation. Uncertainty in fluid temperatures, wall temperatures and mass flows measurements are shown in Table 1, which combined with Equation (3), yield a maximum uncertainty in heat transfer measurements of 8% of the measured magnitude.
equation(3)
View the MathML source