Academia.eduAcademia.edu

Figure 2 – uploaded by Jinwan Kim

See full PDFdownloadDownload figure
Table 2. Lower and upper bound of each design variable and their non-dimensionlization. ables represent the dependent ones that affect a cer- tain response due to the reaction of the design vari- ables.

Table 2 Lower and upper bound of each design variable and their non-dimensionlization. ables represent the dependent ones that affect a cer- tain response due to the reaction of the design vari- ables.

Related Figures (17)

Fig. 1. Fuel cell stack block diagram. For future control purposes, a dynamic PEMFC stack model was developed instead of steady-state model. The dynamic model for the fuel cell stack is comprised of six parts as shown in Fig. 1: (1) mem- brane hydration, (2) cathode mass flow, (3) anode mass flow, (4) stack voltage, (5) cathode GDL, and (6) anode GDL. In this paper, a simplified model was obtained by assuming that the stack temperature is constant. This assumption is justified because the stack temperature changes relatively slowly, com- pared with the 100ms transient dynamics included in the model to be developed. Additionally, it is also assumed that the temperature is perfectly controlled by cooling systems.
Fig. 1. Fuel cell stack block diagram. For future control purposes, a dynamic PEMFC stack model was developed instead of steady-state model. The dynamic model for the fuel cell stack is comprised of six parts as shown in Fig. 1: (1) mem- brane hydration, (2) cathode mass flow, (3) anode mass flow, (4) stack voltage, (5) cathode GDL, and (6) anode GDL. In this paper, a simplified model was obtained by assuming that the stack temperature is constant. This assumption is justified because the stack temperature changes relatively slowly, com- pared with the 100ms transient dynamics included in the model to be developed. Additionally, it is also assumed that the temperature is perfectly controlled by cooling systems.
Using the ideal gas state equation, the partial pressure for oxygen, nitrogen, and vapor inside the cathode channel flow can be represented as follows: Oxygen, nitrogen, and water dynamic characteris- tics inside the cathode can be expressed using the four mass flow conservation equations.
Using the ideal gas state equation, the partial pressure for oxygen, nitrogen, and vapor inside the cathode channel flow can be represented as follows: Oxygen, nitrogen, and water dynamic characteris- tics inside the cathode can be expressed using the four mass flow conservation equations.
The mass flow at the cathode inlet materials can be obtained as The total mass flow at the cathode outlet can be expressed by the equation below [9].
The mass flow at the cathode inlet materials can be obtained as The total mass flow at the cathode outlet can be expressed by the equation below [9].
2.4 Stack voltage model The diffusion coefficient at the electrolyte D,, can be represented by the empirical equation below [20]. Fuel cell stack modeling can be expressed by the term which thermodynamically uses open circuit voltage (OCV), irreversible activity voltage loss, re- sistance voltage loss, and concentration voltage loss.
2.4 Stack voltage model The diffusion coefficient at the electrolyte D,, can be represented by the empirical equation below [20]. Fuel cell stack modeling can be expressed by the term which thermodynamically uses open circuit voltage (OCV), irreversible activity voltage loss, re- sistance voltage loss, and concentration voltage loss.
Table. Parameters used in the simulation [9]. 2.5 Cathode GDL model 2.5.1 Gas model Using the ideal gas state equation, the following equations are obtained:
Table. Parameters used in the simulation [9]. 2.5 Cathode GDL model 2.5.1 Gas model Using the ideal gas state equation, the following equations are obtained:
The number of activated oxygen and produced wa- ter content can be calculated from the electrochemical relation as The water steam partial pressure inside the diffu- sion layers is evaluated as
The number of activated oxygen and produced wa- ter content can be calculated from the electrochemical relation as The water steam partial pressure inside the diffu- sion layers is evaluated as
Fig. 2. Experimental set of the PEMFC system. In this study, experiments are performed to calibrate the dynamic fuel cell model describing the Ballard Nexa. The model was utilized to control the stack
Fig. 2. Experimental set of the PEMFC system. In this study, experiments are performed to calibrate the dynamic fuel cell model describing the Ballard Nexa. The model was utilized to control the stack
Fig. 3. PEMFC model in Matlab/Simulink. In this research, the temperature and pressure of the inputs of anode and cathode are set to a normal state, while the relative humidity of each input is set to zero. The mass flow rate of the cathode input and the compressor output were obtained as follows:
Fig. 3. PEMFC model in Matlab/Simulink. In this research, the temperature and pressure of the inputs of anode and cathode are set to a normal state, while the relative humidity of each input is set to zero. The mass flow rate of the cathode input and the compressor output were obtained as follows:
Each component’s mass flow rate to the membrane via a humidifier is calculated as As the relative humidity at the anode input is zero only the mass flow rate of hydrogen is required as
Each component’s mass flow rate to the membrane via a humidifier is calculated as As the relative humidity at the anode input is zero only the mass flow rate of hydrogen is required as
Fig. 4. I-V Characteristic curve comparison. Fig. 5. I-P Characteristic curve comparison.
Fig. 4. I-V Characteristic curve comparison. Fig. 5. I-P Characteristic curve comparison.
Table 3. Analysis results of the selected experimental points. consumes the energy from PEMFC, the BOP energy consumption should be considered in the optimization process. The maximum consumed energy of BOP is obtained in the experiment as
Table 3. Analysis results of the selected experimental points. consumes the energy from PEMFC, the BOP energy consumption should be considered in the optimization process. The maximum consumed energy of BOP is obtained in the experiment as
Table 4. Coefficients for the constructed response surface of power. Table 5. Description of optimization cases.
Table 4. Coefficients for the constructed response surface of power. Table 5. Description of optimization cases.
Table 6. Optimal results of each case. The BOP power consumption is mainly related with two power demands. One is from the air pump and the other is from the cooling fan. The actual BOP energy consumption can be represented using the weighting factor of air pump and cooling fan, Qa. ,a,, respectively, as
Table 6. Optimal results of each case. The BOP power consumption is mainly related with two power demands. One is from the air pump and the other is from the cooling fan. The actual BOP energy consumption can be represented using the weighting factor of air pump and cooling fan, Qa. ,a,, respectively, as
AS shown in Labie 6, the maximum FEMEFC total power can be obtained by changing the operating variables of stack current, temperature, oxygen and hydrogen excess ratio, and cathode input humidity
AS shown in Labie 6, the maximum FEMEFC total power can be obtained by changing the operating variables of stack current, temperature, oxygen and hydrogen excess ratio, and cathode input humidity
Related topics:
Mechanical EngineeringApplied MathematicsDynamic ModelingPEM fuel cellHydrogen EngineGas Diffusion Layerpower efficiency
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved