Battery pack Introduction
Our approach to design battery pack start in the first place it's to decide which type of cells [] we would like to use for our pack. It's crucial to find the right compromise between some important characteristics for an electric vehicle, because a bad selection could lead to a battery pack that is to much heavy and most of all unsafe. The most significant features for these type of cells are:
- energy density
- C rate
of course it is also possible to define others secondary details such as the type of junction or the physics state of the chemistry present into the cells (Solid or liquid) ecc...
The second step is defining how many kWh do you need for electric retrofit Ex: []
In our case, we have chosen a particular family of lithium battery that is the LiFePo4 battery as this chemistry shows some particular advantages, which is reported below. The chemistry of this battery is lithium based and the reaction that take place involve the reduction of lithium and the oxidation of iron, present as iron phosphate compound. Junction cells have been designed in 1996 (for more detail here: []). Energy density of this type of cells is about 90-120 Wh/kg. It isn't the best density achievable for the family of lithium battery, in fact a medium battery pack of 25 kWh weigh:
25000 (Wh) / 100 (Wh/kg) = 250 kg weight that is not be entirely negligible.
However, with a very short period of time, the energy density could be increased significantly thanks to research as it showed in this available product: which has a density of 140 Wh/kg.
Weight in a electric retrofit conversion is really important because a heavier or not well distribute weight between front and rear axis could cause instability or failure in the car. Our specific prototype had 1105 kg mass with the old engine. After valuation, we decided to introduce an initially battery pack (144 v 180 ah) in which only cells composition weigh:
3,3 kg each cell x 90 cells = 297 kg divided 60 cells front and 30 cells rear
This choice has been taken in order to keep the original distribution of mass and occupy the available volume into the car.
The durability of this cell is about 2000 cycle at 80% of DOD; this means you can discharge until 80% of nominal capacity and the cell life will reach 2000 cycle. A possible example of life span is : 1 cycle of complete discharge and charge each day means that the battery after 6 years still has got the 85 % of capacity. Regarding security Lifepo4 is better secure than other lithium battery, this is a very important feature for vehicle because in case of accident or lost of control the battery pack could be source of dangers .
Our initially battery pack
Our idea of battery packs is composed by 90 LifePo4 cells divided in two volumes, the first one situated inside the engine compartment and the second one situated instead of gas tank. The reason of this choice is to keep a good distribution of weight, in addition to optimize space into the car volume.
Cells type are Lifepo4 90 ah 3.2v prismatic form produced by Winston, more detail here: After a short investigation the free volume in our car provided that is really complicated to introduce all this cells. Anyway only with this configuration is possible to have good performance as well as gas engine, trying to reach 200-250 km of range.
Our first attempt in order to combine free space into the car and battery pack appeared difficult and full of hidden tricks, as cars have not square shapes. In fact, to simplify connection and minimize dispersion trough wires, battery pack must have rectangular or similar shapes that are completely different from cars volume.
The first approach has been to sketch outline with a cad software but it was immediately clear that was a difficult way to reproduce an engine compartment, and for this reason we decided to build cardboard volume and approach problems in the simplest way as possible. Parallelepiped cardboards are not complicated to build and they appear an initially good and simple solution to fit our battery volumes into the car. The problem after few attempts was that cardboard adapt itself excessively to corners and did not copy the reality.
Discarded that idea we tried to resolve the problem with a entry level 3d scanner provided by a member of our association . The scanner did not provide good results, we didn't know if the engine compartment was not well illuminated or too dirty, so after a day of experiments we decided to stop scanning. The used 3D scanner may be not the best product in commerce but it's really incapable to scan surface also with a bad resolution. However, in the future this technology should be more investigated to discover if this technology is good enough to replicate big and low reflective objects. This way may also permit to draw and re design volume, and fit a battery pack in a 3D car assemble.
However we decided to improve parallelepipeds volumes and build chipboard ones; it appears immediately more detailed and many little problems are better showed with rigid structure. This strategy could seem simple but it require experimentation. Further more it is more expensive than cardboard but it offer better results.
Until today we built a first front aluminium battery which permit to place 45 cells, reaching the voltage of 144v and 90ah capacity; this battery is useful for first engine start and initially setting. We are going to design rear pack necessary to balance weight and useful to maximize range and electric power energy absorption .
First front battery pack
Our first front battery pack is composed of 45 cells in series, which are necessary to reach 144v dc. Every cell has 3.2v nominal and a range between 2,65 v (when discharge) to 3,60 v (when fully charge). The maximum power of this battery pack must be calculated by the sum of nominal voltage 144 v and the maximum continuous ampere available for every single cell.
Our 90 ah LiFePo4 can reach 3C of continuous discharge; in this way the total ampere available are 3x90 ah and 270 A can be continuously drained from our battery. The total power is 270 A plus 144 v = 38800 W. Although with this maximum power we can't supply the fully power of the motor, it's not a problem because our controller  is able to limit the current absorption to a specific set value.
Our 45 cells are organized in three lines, each composed of 14 cells, and in a fourth line made of 3 cells that are rotated by 90°. This disposition derive from thinking about space volume and due to connection cabling have been previously showed. The external case (dimensions 860x500x250 mm) has been built with 30/10 mm aluminum sheet and a plexiglas top. Over the top there is the main junction box useful for main electric power connection fuses and main switch. This assembly must be develop to reach more strong case and major security case, but for our first test and it's good enough.
Aluminum case is lighter than steel but is more complicate to weld and assemble. Another positive characteristic of aluminum is the corrosion strength , in fact it's auto passivating material, so they don't need any superficial protecting treatment. Opposed to this advantages there is a lower mechanical property compared to steel, so major thickness is needed to replace steel with the same strength.
We are going to develop future features such as :
- battery case comply whit automotive normatives
- different system of sheet metal housing more strong and secure
- develop a circuit power board necessary to integrate relays and minor security fuses
- develop rear battery pack to improve autonomy and maximum power
- develop a heating pad system to provide battery pack from under zero degree temperature
- improve connection with slave balancing board to simplify wiring.
To write this document we referred to the following Internet shared resources: