F+L Week 2018 | Robert Wardle | Investigation of Diesel Injector Nozzle Deposit Accumulation Mechanism Using High FAME Content Biodiesel
An engine test was conducted under conditions that favored nozzle hole deposit formation using 20% palm oil methyl ester biodiesel (B20). After the test, the composition and quantity of the deposit formed was analyzed in order to elucidate the deposit accumulation mechanism and its effect on engine performance. Moreover, for further understanding of this mechanism, the engine test and deposit analysis were conducted using the same B20 fuel treated with a DCA (Deposit Control Additive) which functions as a surfactant and a dispersant. Concerning the test method, in order to create a severe condition for deposit formation and accumulation, firstly the operating pattern of the engine test was designed to repeat the full load condition for eight hours and the soak condition for one hour, which increases the temperature inside the injectors. Secondly, the fuel return temperature was increased by raising the common rail pressure. The hot return fuel was introduced into a buffer tank set between the engine and the fresh fuel line giving a longer residence time for fuel oxidation. Thirdly, zinc neodecanoate (2mg/kg of Zn) was added to accelerate the deposit formation. During the tests, injectors were replaced at 100 hrs in order to confirm the change in the deposits over time in terms of structure, composition and quantity, with the aim of elucidating the deposit accumulation mechanism. Figure 1 shows the SEM image of the deposit at the inlet and outlet when running on B20. From Figure 1, it is obvious that more deposit was formed at the outlet as compared to the inlet and the deposit increased with time especially at the outlet. These results were confirmed by the measurement of deposit thickness by a laser microscope. Moreover, the main component of the deposit was zinc carboxylate both at the inlet and the outlet of the nozzle hole, which was demonstrated by a combination of the results of EDX elemental analysis and micro-FTIR. From the aforementioned results, it is inferred that the deposit accumulation mechanism follows the steps shown in Figure 2. At step 1, polar oxidized fuel adheres to the nozzle wall, forming a boundary layer composed of zinc carboxylate on which the growth of the deposit is favored, which then forms a thicker deposit layer (step 2). The deposit cracks due to the difference in thermal expansion coefficients between the metal injector and the deposit (step 3), after which the fuel trapped in the cracks (step 4) becomes oxidized (step 5) and forms secondary deposit in the cracks (step 6). Similar to step 3, at step 7 secondary deposit cracking occurs and all of these steps repeat before the deposit accumulation eventually reaches an equilibrium state where the formation and the removal of the deposit become equal. In the next test, in order to further understand the deposit accumulation mechanism, the same engine test was conducted using B20+DCA (B20 treated with Deposit Control Additive, again with Zn added). The changes in the deposit structure, composition and quantity were examined by comparing with the B20 case. The results indicated that the DCA behaved both as a surfactant, delaying the formation of the boundary layer, and as a dispersant, inhibiting the agglomeration of the deposit, which eventually hindered accumulation. Finally, because the formation and accumulation of the deposit may impact upon engine performance, the effect of the deposit on the torque loss was investigated and the results are shown in Figure 3. In the case of B20, the torque dropped by approximately 10% after 50 hours. After that, although the deposit became thicker, the torque change was small which indicates that apart from deposit thickness other factors including surface structure also affect engine performance. By comparison, in the case of B20+DCA, because of the surfactant and dispersant function of the DCA, which led to a slower boundary layer formation and deposit growth, the torque loss started at 150 hours and resulted in a final reduction of 3%. These results are in good agreement with the deposit accumulation mechanism.