One of the hardest part (at least for us) of building an electric motorcycle is the development of the accumulator and respective container. The sizing of the battery is not the only problem, designing a safe and simple wiring method, and containers for the batteries that comply with the safety norms of Motostudent regulation, and at the same time actually fit all of that plus electronics inside the frame of a motorcycle is quite difficult as well. This is the story of how it all started, the sizing, testing and test bench assembly of the TLM02e battery pack.
Before we get started, we have to remember that this is the team's first electric motorcycle, and second ever motorcycle. This means absolutely no experience on this matter and thus lots of rookie mistakes, but it's just part of the process.
The electric system for the motorcycle, including the battery pack was designed and built as a Masters Thesis of one of the members.
The first step to size the battery was to create a motorcycle model in order to have a rough idea of the amount of energy and power required to go through the final race in the Aragón Motorland GP international circuit.
To create a simple model of the motorcycle, we opted for a 3 block approach: Pilot block, Electric System block and Motorcycle Dynamics block.
The Pilot block, as the name suggests, simulates the throttle and braking behaviour of the pilot, that are determined from speed and position on the track. The throttle and brake outputs are then introduced into the motor model, which includes the limitations and efficiencies of the motor and controller and the required energy is determined. The Motorcycle Dynamics block takes the motor torque output and introduces into a simple single mass model in Simulink, thus obtaining acceleration, speed and position, which are looped back into the pilot model.
Testing this configuration to obtain minimum race time yielded a front to back sprocket ratio of aproximately 0.2 and around 4.3 kWh of energy.
The images below shows the simulated motorcycle and pilot variables for the final race.
Having the required energy it was time to choose a cell chemistry and pack configuration.
From the competition regulations, the maximum permitted voltage for the battery pack is 110V which limits the number of series connected cells (generally lithium chemistries have maximum and minimum voltages of around 4.2V and 2.5V) to 27 cells (if charged to 4.07 V maximum).
Regarding the number of parallel connections, it depends on the energy density of the chosen cells, thus began our quest to find the most suited cells to our design performance and cost-wise.
After an intensive search for the batteries with the best performance, both as unitary cells or as a battery pack, a list of the candidates was organized in the following table:
Starting to consider the options by removing the last two candidates for economic reasons, even though these are the only renowed brands in this table and these offer the advantage of high cycle life and simplified battery pack thanks to the high capacity single cells, their Wh/Eur figure of merit is much lower than the other ones.
At this point multistar, turnigy and westart cells remain. The problem is that no accurate distinction between these can be made because their manufacturer details are obtained at very different testing conditions. For example, Turnigy cells claim their capacity for their nominal 20C rating while Westart does it only for 1C. Then a direct comparison would be unfair for the Turnigy cells given that the higher the discharge rate the lower is the energy provided by the cells because of their internal resistance.
At this point, there was some skepticism on the last battery cells manufacturer data and no direct comparison could be made, for the reasons mentioned above.
In this way, an automated experimental system was designed to aid in testing cells under similar conditions and have a fair way of comparison.
As we don't want to make this post too extensive, a segment just about the automated cell tester will be posted at another time.
The experimental setup for the discharge tests can be seen below.
Taking into account the estimated minimum energy capacity of 4.276 kWh and a nominal battery voltage of about 100V (maximum permitted is 110V by regulation), in a rough calculation the minimum battery charge capacity needed is 42.76Ah but 50Ah is considered as safety margin. Considering the nominal motor current of 153 A the minimum C rating requested to the batteries will be 3C. In this way, the cell tests are conducted at this rating.
The discharge curve and the numerical results for the tests conducted on the various cells is shown below:
Taking these results into account and adding to this the Wh/eur figure of merit it can be concluded that the best option are the Multistar cells from the Multistar battery pack.
The main disadvantage of these cells is the fact that it would be necessary to have many of these in parallel for a 50Ah+ battery pack. Trying to attenuate this problem a higher capacity 16Ah multistar cell is used, this cell was separately tested and virtually the same energy density was obtained and the new cost figure of merit improved to 5.32 [Wh/eur].
As said before, it was chosen to connect 27 cells in series and only charging them to 4.07 volts. This means we are actually only using about 86% of the total energy of the battery.
Turning to the number of parallel cells, there is, for a 3p and for a 4p setup, a capacity of 4.265 kWh and 5.688 kWh respectively. In this way a 4p setup is used yielding a good safety margin at the cost of more 8Kg of weight.
Now for the parallel strings there are several options regarding their interconnections as is illustrated in the image below. On option (a) more BMS (Battery Management System) boards would be necessary since parallel group of cells could have different voltages. Since this represents a bigger cost and complexity in terms of BMS boards option (b) is adopted were parallel cells are directly interconnected, this saves BMS boards but preliminary care must be taken about the currents that cross these parallel interconnections, in terms of the current carrying capabilities and cell voltage reading errors on the BMS.
To get an idea of how parallel cells behave when it comes to current distribution, two experiences were held. First experience consisted in discharging four parallel cells through a resistance and monitoring the current passing through each of them, having obtained the results shown in the image below:
It can be seen that the cell 2 has a higher internal resistance because it was the one that discharged more slowly, but in the end, the situation reverts since the cell 2 tries to compensate for the final of SOC’s voltage drop of the other cells.
This was be a reason for having to purchase more cells than are needed in order to do a selection of cells with similar internal resistance for use in parallel. In addition, since the biggest imbalances happen at the end of the discharge, having a bigger capacity battery pack helps to avoid this final imbalance, avoiding necessity of much more robust parallel connections.
The second experience consisted in discharging a 2s2p configuration of cells and monitoring i(t) as presented in the images below, together with the experience result on the right. This experiment was carried out at 3C.
It can be seen that the maximum current approaches 5A again at the end of the discharge. With two more parallel cells (4p total) in the worst case there would be more 3 times the actual current imbalance totaling a maximum of 15A in the most stressed interconnection. This is an important factor when sizing the connectors and conductors for the parallel connections between cells. All the cells were then tested so that their internal resistance could be determined and their configuration in the pack was chosen by connecting the the extremes (internal resistance wise) in parallel.
Having defined all the configurations, it was time to decide how to fit all of this into a motorcycle and how to connect it all. One of the worst mistakes we made was to design the motorcycle frame without really thinking the battery configuration through. So, we had to adapt it to the frame. The idea was to split the battery pack into 2 modules and place them above each other in the frame. The first proposal for the placement of the various parts of the electric system is shown below.
The function of each of these elements will be explained in other posts.
After the removal of the original package a balancing connector (TE 1-480318-0)was soldered to the terminals of each cell and XT150 power connectors were soldered to the terminals of the pack as shown below. The soldering process had to be done carefully to avoid overheating the terminals and therefore possible damage to the cell.
After soldering all cell terminals the balancing and BMS wire harness was built, as shown below. The connections to the BMS (blue wire) were twisted in order to decrease the EMI absorption and consequently noise in cell voltage reading.
All the wires used in the battery pack were made of silicone which is high temperature resistant (+-180 oC versus traditional 105oC) and non-flammable making these compliant with motostudent regulations. The current rating also played a role in the choice of wire conductor area in which the maximum current the blue wires had to withstand was the balancing current of circa 7.5 [A] and circa 15 [A] for the grey wires of the parallel cell interconnections.
PLA cell spacers (white) were printed to hold the cells together while keeping a small space between them for the air to flow through. These spacers were also meant to keep the cells still inside an aluminum frame that would be put inside the final container and screwed to the motorcycle frame.
The temperature sensors used are based on NTC resistors (NCP18XH103F03RB) with SMD
encapsulation manually soldered to wires connecting to BMS multiplexers. To avoid short circuits and damages in the cells these sensors were covered in epoxy glue near the NTC zone and latter with electrical tape for extra safety.
The final battery pack and harness with the BMS connections can be seen below.
The main objectives of test bench were to test the thermal evoulution of the motor, motor controller and batteries, monitor the energy consumed, thus obtaining the total energy contained in the final battery pack, test the quality of the power connections and finally test the automatic shutdown system.
In order to meet the objectives the experiment shown below was prepared.
This experiment consisted of a test bench in which the motor of the motorcycle was coupled through a transmission chain to a generator that generated electric energy for a resistive load of 12.3 kW. However due to the maximum power of the generator the dissipated power was limited to 10 kW.
The manufacturer does not indicate the cell working temperature range, only the maximum ambient temperature at which they can work (4-49 ° C). For this reason, additional literature concerning batteries of the same technology had to be consulted to determine the most common limits of the operating temperature of these cells. The average statements found in the bibliography claimed the range of -20 to 60 °C, which was chosen as the limit for cells.
In order to try to match the conditions that will be found in the motorcycle or that are required by the motor manufacturer, several fans of various types were used for the motor controller, motor and battery. It should be kept in mind that the measured airflow of the forced ventilation is in all cases much smaller than the one that would be found on a motorcycle.
The referigeration points can be seen below.
Thanks to the built electronics and motor controller’s CAN network compatibility all the necessary data was available from the test bench and acquired with a USB to CAN device. The entire acquisition scheme is shown below, here the USB to CAN device and its software “IXXAT MiniMon” saves the CAN data to a *.cvs type file which is then opened via a purpose written matlab script that enabled to view graphics of the variables during the course of the experiment.
During the experiment, ajustments had to be made in order to keep the power at the generator nominal point, as the motor torqque decreases as the battery voltage decreases. The mechanical data and throttle input throughout the eperiment can be seen below.
The electrical values during thhe experiment can be seen below
Here it is confirmed that the end of the experiment happened due to the automatic shutdown system, since the minimum voltage of a cell got
inferior to the configured 3 volts.
In the following graphics are represented the power and energy evolution with time.
The obtained 5.5kWh energy capacity comes close to the 5.6kWh expected in the estimations.
Finally, the cells, motor and controlller temperatures is shown below.
The fact that no element surpassed it's temperature limit is indicative that if simimlar conditions ,e.g. ambient temperature, are met, the system will be able to deliver more power.
Thermal imaging also reveals the hotest spots and maximum temperatures reached in the system.
In conclusion, the testbench experiments were satisfatory and corresponded to the theorectical calculations. This first step was very important as it allowed us to test the behaviour of batteries and obtain data that will certainly prove useful in further projects and eperiments. Having this design validated, the next step was to incorporate it in the motorcycle. Various changes had to be made in order to have a structure that would fit in the motorcycle frame, although the general idea is the one that was presented here. In the next post, we will go through the changes and construction of the final battery pack for the TLM02e Prototype.
Hope you enjoyed reading this post and are open to suggestions and comments. We also want to thank Farnell Element 14 once again for supporting our project!