Project 2 Thermal modeling of battery pack

Aim :

For a 10 cell series lithium ion battery model, simulate the thermal effects and compare life cycle performance at various temperatures, charge & discharge rates using MATLAB.

Simulink model :

First, we used the table-based battery, then we instrumented it, added a thermal port, and left the rest of the parameters at their default values. We used default data from Matlab rather than data from any data sheet. And after that, we connected 10 of these cells in a row, all with the same setup. As a result, the pack's voltage is multiplied by 10 while maintaining a constant capacity. In order to take the output voltage from the voltage sensor and receive the voltage signal, a goto block is connected to the voltage sensor via a ps to simulink converter. A from block is then connected to the scope. A current sensor is connected to the controlled current source, which is connected to the battery to draw current from it. The current is then connected to the block through a ps to simulink converter, where a from block is present to receive the signal and transmit it to the scope. The same current is now integrated through an integrator block and divided by the pack's total capacity, 27Ah (27*3600), to produce an ampere second, which is then subtracted from one to produce the battery's SOC and a from block is connected to the scope to receive the SOC signal, and a goto block is connected to take the output SOC. UDDS drive cycle data is employed in the form of current with regard to time and a signal builder is used to provide the input to the controlled current sensor. In order to increase the area of contact and as we provide the temperature input to the temperature source block, the temperature source is now connected to the thermal port to provide the external temperature. This connection is made through the convective heat transfer block rather than directly. In order to gauge the battery pack's temperature, a temperature sensor is connected there, and one port is connected to a thermal reference. The temperature signal is sent from the temperature sensor's output through a ps to simulink converter block and a from block that is linked to the scope. Finally, a solver configuration and an electrical reference are attached to the model's electrical side for ground. As the input current is only for the first 1369 seconds, we run the simulation for that length of time because after that, the graph will remain unchanged because no load current is being drawn from it.

Result :

The blocks are all given default values at this place, and the simulation results are shown. Because we used it the same way, the current drawan is shown in the first graph. You can see that the greatest value is up to 8 and the smallest value is -5, which represents charging regen braking. The second graph depicts a SOC that is falling yet linear; depending on whether the current is positive or negative, it is occasionally charging and occasionally discharging. After running the entire simulation, it is at 0.992, which is approximately full, starting from 1 which is full charge. The third graph shows the temperature, which is the default graph for which I made no adjustments. As a result, it rose marginally from 298.1 to 298.5. The voltage fluctuation is the final to be discussed. It begins at 41.9 volts and concludes at the same 48.9 volts throughout time, growing and decreasing as the battery is charged or discharged or as per the current drawn from or supplied. 42.2 volts at maximum draw and 41.2 volts at minimum draw.

We raise the temperature here to 350k from the temperature source. Additionally, we made an area from 1e-1 to 1m2 for convective heat transfer. Let's look at the outcome; the simulation time and all other parameters are the same.

Because we didn't make any modifications to the load side, the first graph of current is the same as what we can see above. Similar to the previous graph. The runtime is so low that we cannot see any differences in the second graph, which represents the social sector. The third graph, which shows temperature as a function of time, has been altered so that it first starts at 300, grows to 350, and then remains constant for 200 seconds. So yeah, there is a difference between this one and the previous one. The voltage graph comes last and is remarkably identical to the one before it.

By changing the temperature parameter once more to be negative to see what happens. Additionally, the area of convective heat transfer was 50 m2, and the coefficient was 1.

Due to no modifications being made on the load side, the current is the same as it was above. Since the running time is so low, we cannot see any differences in the second graph, which is a SOC, as I previously mentioned. The third graph, which shows temperature as a function of time, has been altered similarly to the first and second ones. It starts at 300 and decreases until 200, at which point it becomes constant. So, yes, we have noticed the stark contrast between this one and the previous one.

Conclusion :

The cycle of charge and discharge is not very variable here due to temperature differences, but in real life it must be variable and the temperature of the battery changes by varying the outside temperature supplied.