Accordingly it is indicated that the maximum voltage of switch of Vsmax is suppressed to less than 1.2 times of input voltage.
Fig. 10 and Fig. 11 show the distribution of the maximum current of switch of Ismax* as switching stress. In the same manner as Vsmax*, Ismax* is suppressed when μ is less than μ=3.0.
From Fig. 8 〜 Fig. 11, it is shown that the sweitching stress is approximately constant in spite of changing λ and is depending on the value of μ. And the switching stress is suppressed when the value ofμ is less than μ=3.0. Moreover the switching stress does not exceed the value of circuit design in order that μ becomes low according to reducing instantly the value of Lo at Curie temperature.
Fig. 12 and Fig. 13 show the distribution of the normalized output power of Po*. In these figures, Po* becomes high according to enlarging μ and λ. However the value of λ has upper limit in order that the resistance of induction heating load is small.
From the above considerations, μ=2.7 and λ=0.6 are selected for circuit de sign as the values of the normalized frequency of μ and the normalized resistance of λ, respectively. The specification of power supply is shown in Table 2.
Fig. 14 Outline of experimental system
Fig. 15 Heating state
Fig. 16 Structure of metallic filter
5. EXPERIMENTAL RESULTS
5. Experimental system
In order to prove the reduction of PM the experiment by PM reduction system using high frequency induction heating is performed based on numerical analysis.
Fig. 14 shows the outline of experimental system. In Fig. 14, exhaust gas that divided from exhaust pipe is led to dilution tunnel through induction heating unit. And the pipe that bypasses induction heating unit is prepared in order to compare the effect of PM reduction system. As for measurement of PM the dilution method is used.
The marine diesel engine, which is 4 cycles and 3 cylinders, is used for this experiment. The maximum output of this engine is 100[PS] and fuel oil A is used. PM is measured on three conditions as follows.
(A) Exhaust gas bypasses PM reduction system.
(B) Insufficient heating state of metallic filter. (700K)
(C) Sufficient heating state of metallic filter at high temperature. (1200K)
Fig. 15 shows the heating state of metallic filter at high temperature as condition (C). The temperature of metallic filter is approximately 1200 [K]. In this state, PM is burned instantly and is reduced.
Fig. 16 illustrates the structure of metallic filter shown in Fig. 15. This filter is constituted by the bundled stainless steel tube of diameter 10[mm]. As illustrated in Fig. 16, the upper cover and lower cover are set to the circumference part of the gas entrance and the gas exit. And in the central part of filter, the stopper is set to middle of tube. Therefore, exhaust gas flows to direction of the arrow that illustrated in Fig. 16. As the results, in order that residence time of exhaust gas in filter is long, PM is burned efficiently.
5.2 Effects of PM reduction system
PM is measured by dilution method in order to prove the effects of PM reduction system. Table 3 indicated the effects of PM reduction system. Trapped PM in the filter of dilution tunnel is shown in Fig. 17. Filter numbers in Table 3 and Fig. 17 are linked. In Table 3, the density of PM in exhaust gas at condition (C) is clearly low in comparison with the density of PM at condition (A) and condition (B).
