MARINE DIESEL ENGINE PISTON S INSTANTANEOUS POSITION EXPERIMENTAL DETERMINATION

Transcription

1 PROBLEMY EKSPLOATACJI 113 Przemysław KOWALAK Akademia Morska w Szczecinie MARINE DIESEL ENGINE PISTON S INSTANTANEOUS POSITION EXPERIMENTAL DETERMINATION Key words Marine diesel engines, diesel engines indication, piston TDC determination. Summary The experimental possibility of piston movement observation through the crankcase has been presented. The characteristic piston positioning points in the time domain have been determined. The piston TDC estimation methodology based on observed positioning points has been developed. A mathematic model for piston TDC estimation has been proposed. The difference between thermodynamic and dynamic TDC has been detected and presented. Introduction Several methods for piston s position determination have been developed recently. Most of the work is concerned with the piston s TDC detection [1], [3], [4], [5]. The TDC markers installed on the engine s flying wheel are known as the easiest way to achieve that goal. Due to the static method of calibration, the TDC determined in this way is known as static TDC [5]. There are known other methods too, based on compression curve analysis [1]. Assuming that the compression is an adiabatic process and the charged air mass loss does not exist, the peak pressure should occur in geometrical piston TDC. Unfortunately, due to heat transfer and mass losses, a shift exist between the ideal and the measured pressure peak [3]. The TDC detected by means of compression analysis is known as thermody-

2 114 PROBLEMY EKSPLOATACJI namic TDC [3], [3]. In some laboratories, a cylinder head modification can be made for proximity sensor installation and direct piston position detection in the neighborhood of the TDC [5]. As a result, a dynamic TDC is obtained which is recognized as the closest to the real TDC [5]. The total costs of the TDC determination usually increase with the accuracy. Using a static TDC is a relatively cheap and uncomplicated. However, it does not take into account the shaft torsion due to the high torque that can be close to 10 7 Nm in marine long stroke engines. The torsional vibrations, the movement of a journal in a bearing shell and crank arm bending are omitted too. Determination of the thermodynamic TDC [3], [5] does not require additional measuring equipment but is sensitive to heat transfer and mass loss estimation. This especially concerns low speed marine engines having large cooling surfaces. Additional error can be generated due to a deterioration of the pressure curve, since the pressure is sensed via long and narrow bores in the cylinder head in marine engines. Installation of the proximity sensors is expensive and requires cylinder head modification, which can be impossible or banned by classification societies. 1. The experiment goal Proper TDC positioning on the indicator diagram as well as the pressure cylinder volume relation are the main factors affecting the mean indication pressure (MIP) estimation accuracy. Since the problem is still not fully solved, further research on the piston positioning of a working engine is needed. The target of the experiment was the determination of the real piston course in the working engine based on its instantaneous position observation. Furthermore, the trial of dynamic TDC estimation was the object of interest. The objective is to develop a methodology that could allow the verification of the analytic model with an experimental result. For comparison purposes, the pressure in the cylinder was simultaneously recorded. To enable the thermodynamic TDC determination, as defined in [3], the experiment was performed with the fuel cut off to the relevant cylinder. 2. Experiment details The experiment was made on an eight-cylinder, four-stroke CI engine Buckau Wolf type 8RDV136. The engine is loaded with a fixed screw propeller submerged in a water pool. Only the third cylinder, when counting from the flying wheel side, was observed. A measuring assembly designed for piston positioning was developed [2]. Its principle is based on the amount of light reflected from the reflective rod attached to the piston skirt. Since the rod has a cut toothed pattern, the amount of light varies periodically when the piston moves.

3 PROBLEMY EKSPLOATACJI 115 For piston movement observation, an optic head type FU-85Z by Keyence was used. In the cylinder, pressure was measured by means of an optoelectronic Optrand sensor type F532A8-ACu. Due to construction restrictions, it was impossible to prepare the reflective rod for the entire piston stroke observation. A satisfactory signal was obtained in the range of mm below TDC, which corresponds to degrees of crank angle (ºCA). The resolution of piston position is 6mm. Finally, 42 upper positioning points were chosen for the analysis. Fig. 1. Measurement system diagram for the piston s positioning; 1 optical sensor, 2 reflective template, P optoelectronic transducer, C computer with acquisition card 3. Data analysis methodology The piston positioning signal was processed in such a way that, for every positioning point P n, two coordinates were assigned. One coordinate for geometrical height of the piston above BDC h [m] and one coordinate for the time when point P n was recorded t [s]. The absolute time scale commenced at beginning of the measurement was used. Subsequently, for every positioning point P n, an index was assigned starting from n=1 for the first point P n, recorded when the piston moves upwards and ending with n=84 for the last point recorded when the piston moves downwards. During one revolution of the crankshaft, a set of points P 1 to P 84 was acquired. The measured data from two crankshaft s revolutions are presented on Fig. 1. Assuming that the engine speed is constant and uniform, the time when the piston passes the TDC can be derived from the following formula: t84 ( n 1) tn t TDCi = + t 2 n for i 1, n (1) 2 For uniform engine speed, calculated TDCi t coordinates for estimated TDC points P TDCi should have the same value regardless from which piston position n they were derived.

4 116 PROBLEMY EKSPLOATACJI Fig. 2. Sample of processed piston positioning signal To allow comparison results from consecutive engine cycles, the absolute t TDCi values were recalculated into values relative to their mean value for one considered crankshaft rotation as follows: ttdci i T TDCi = ttdci 42 (2) Because the TDC position is usually given in CA, the conversion from time domain into angle domain was made by the simplifying assumption of constant angular speed of the engine shaft ω as follows: 360 ω TTDCi α CAi = (3) 2π The absolute time of thermodynamic TDC ( τ thermo ), which was derived from the pressure curve, was converted into the angle domain in a similar way, based on equations (2) and (3): i 42 = ttdci ω i= τ thermo 42 α = (4) 2π Finally, the angular distance between thermodynamic TDC and the calculated, estimated dynamic TDC positions α CAi was derived. The α was set as the reference point as follows: i 42 = =1

5 PROBLEMY EKSPLOATACJI 117 α RELi = α α (5) CAi In the following step, the piston position h above BDC was set as an argument of function, and the relative angular TDC position α was set as dependent variable. To allow dynamic TDC determination, an approximating function ( h) developed. By means of the least square error method, the function s ) α ( h) were calculated as follows: 2 ( h ) = a h + b h + c RELi α ) was factors α ) (6) Conclusions The above analysis allowed the determination of whether the analysed TDC was noticed at motored or working revolution of a four-stroke engine. The estimated TDC points form curves of different shapes for motored and working revolution (Fig. 2). After thirty consecutive revolutions had been analysed, high repeatability was observed. Since the TDC in the working revolution is of the more interest than the one in the motored revolution, a comparison of 15 consecutive working cycles are presented in the diagram in Fig. 3. Fig. 3. Determined, estimated motored and working TDC positions The course of the determined, estimated TDC versus the corresponding thermodynamic TDC is presented in relation to the mean value of the estimated TDC as per formula (3) and (4). It can be observed that the higher the estimated TDC is observed above BDC the closer it is to the thermodynamic TDC. The thesis can be formulated that the estimated TDC obtained in the described way approaches the real dynamic TDC.

6 118 PROBLEMY EKSPLOATACJI Fig. 4. Thermodynamic TDC α versus estimated TDC α CAi in relation to the mean value Based on the above thesis an approximation function ) α ( h) was used to determine the real dynamic TDC of the piston. Since the piston stroke is 0.36m the value of the function ) α ( 0.36) should give the angular position of the dynamic TDC. The results for a sample of 3 revolutions are presented on Fig. 4. It is worth noticing that the approximating function matches the experimental results to a high degree. The angular positions of the determined dynamic TDC calculated from the formula (6) were in the range of CA. It has to be pointed out that the values are relative to the determined thermodynamic TDC values. The positive values should be understood as the thermodynamic TDC was advanced to the dynamic one. Fig. 5. Dynamic TDC determined by means of approximation function

7 PROBLEMY EKSPLOATACJI 119 Nomenclature n positioning points index, n t n [s] time coordinate of piston positioning points P n t TDCi [s] absolute estimated time of piston TDC passing T TDCi [s] relative estimated time of piston TDC passing ω [1/s] crankshaft mean angular speed of 1 revolution α [deg] relative estimated angle of piston TDC passing CAi τ thermo [s] absolute time of thermodynamic TDC α [deg] relative thermodynamic TDC given in CA degrees ) α RELi [deg] estimated TDC in relation to thermodynamic TDC α ( h) approximating function a, b, c approximating function coefficients h [m] Piston position above BDC References 1. Gałecki Wojciech, Tomczak Leonard, Przenośne systemy diagnostyczne dla procesu spalania w okrętowych silnikach wysokoprężnych, Gdynia 1996, XVIII Międzynarodowe Sympozjum Siłowni Okrętowych, s Kowalak P., Borkowski T., Instantaneous piston speed measurement for marine diesel engine combustion monitoring. International Scientific Conference On Combustion Engines. Journal of KONES 2006, No. 2, Warsaw 2006, vol. 13, p Tazerout M., Le Corre O., Rousseau S. TDC Determination in IC Engines Based on the Thermodynamic Analysis of the Temperature-Entropy Diagram, SAE , 1-10, ISBN Wang Y., Lim T. C., An analysis of modal damping sources in reciprocating engine, Journal of Sound and Vibration, 242 (1), p , Wimmer A., Glaser J., Indykowanie silnika, Warszawa: Instytut Zastosowań Techniki, rozdział 9, s Reviewer: Andrzej BALCERSKI

8 120 PROBLEMY EKSPLOATACJI Eksperymentalne wyznaczanie chwilowego położenia tłoka silnika okrętowego Słowa kluczowe Okrętowe silniki spalinowe, szacowanie położenia tłoka TDC. Streszczenie Przedstawiono eksperymentalne możliwości obserwacji ruchu tłoka poprzez skrzynię korbową silnika okrętowego. Wyznaczono charakterystyczne punkty położenia tłoka w czasie. opracowano metodę estymacji wystąpienia GMP tłoka na podstawie jego obserwowanych położeń. Zaproponowano model matematyczny umożliwiający szacowanie położenia dynamicznego GMP tłoka na podstawie obserwacji jego ruchu. Wykazano występowanie różnicy pomiędzy wystąpieniem GMP termodynamicznym, a GMP dynamicznym.