**1. Introduction**

The performance of the lithium-ion battery (LIB) is at the core of the driving system of an electric vehicle, and thus it significantly a ffects the driving range and service life of electric vehicles [1,2]. Among the many factors that determine battery performance, the influence of temperature on the battery should not be underestimated [3]. Battery heating is an inevitable phenomenon and a complicated problem [4]. Reasonable control of temperature changes in battery operation depends on an e fficient battery thermal managemen<sup>t</sup> system (BTMS) [5,6]. Essentially, batteries work to convert chemical energy into electrical energy for the machine to use, so the heat generated is waste energy, which reduces the energy conversion e fficiency of the battery. A certain amount of heat yield is beneficial to ensure the normal chemical reaction temperature of the battery and promote the charging and discharging process of the battery. However, if an abnormal amount of heat is generated or the heat does not dissipate in time, the battery su ffers overheating, which may potentially cause the active material on the electrodes to peel o ff and promote electrolyte degradation, that is, cause harm to the battery itself [7–9]. These changes are irreversible and can cause permanent damage to the battery. In addition, because the battery is hot, there are safety risks. An overheated battery can short-circuit due to damage to the internal materials, such as the diaphragm, which sets o ff a chain reaction that can lead to a fire or even an explosion [10]. Without e ffective control and management, the harmful effects of a single battery can spread throughout the entire battery pack, and thermal runaway will be amplified, causing the battery pack to be out of control. Therefore, it is very necessary for electric vehicles to design a practical and e fficient BTMS [11]. Excellent thermal managemen<sup>t</sup> benefits the performance of the battery pack [12].

There are three main types of LIBs used in electric vehicles (EVs): (1) prismatic; (2) pouch; and (3) cylindrical batteries. Figure 1 shows these three types of batteries. The prismatic battery can be designed according to the needs of customers. This makes the prismatic battery suitable for almost all kinds of electric cars. This high adaptability also results in prismatic batteries that vary in size, nominal voltage, and other parameters, which makes it di fficult to form an industry standard. Pouch batteries are manufactured and packed by superposition. Compared with prismatic batteries, the aluminum alloy shell of the pouch battery pack is replaced with a lighter aluminum plastic film package, which improves the energy density of the whole battery pack. However, a big drawback of soft-pack batteries is their poor consistency, which means they need more sophisticated control and monitoring systems. After years of development, the cylindrical battery has obtained a high degree of standardization, which makes it easy to achieve a unified industry standard. In addition, the cylindrical battery has inherent advantages in regard to heat dissipation, and a good heat dissipation space is formed between the cylinders when packing.

**Figure 1.** A prismatic battery, pouch battery and cylindrical battery (from left to right).

Among these types of battery, the cylindrical battery is the most discussed and well-studied and it is the first mass-produced commercial battery., Thus, much of the research has focused on cylindrical batteries. Most of the studies are based on three di fferent ways of heat dissipation: AC, liquid-cooling (LC), and phase-change material (PCM) cooling. Wang detailed the three cooling methods in his research [13]. AC is widely used for its simplicity, low cost and there is no hidden danger of battery damage. Wang verified the e ffectiveness of forced-air cooling in ensuring the battery pack operates within the normal temperature range (no more than 40 ◦C) when the discharge of the current rate of a single battery is set at no more than 3C. Similarly, based on the AC method, Mahamud et al. [14] studied the influence of reciprocating airflow in the BTMS of cylindrical battery packs. This method reduced the battery temperature by about 4 ◦C (72%) and the maximum temperature (MaxT) by about 1.5 ◦C. LC is a cooling method that uses a liquid material with a large specific heat capacity (SHC), such as water, to flow through the surface of the battery and take away the generated heat. This method requires the battery system to be highly sealed, but its strong heat transfer capacity leads to a better cooling e ffect than that of AC [15]. Wang et al. [16] proved that several factors, such as fluid flow, flow direction, etc. determine the cooling e ffect at a certain degree. They designed a BTMS based on a hot silicon plate and used experiments and simulations to explore and verify the significant influence of liquid flow, flow direction and the number of cooling channels on the cooling capacity of the BTMS. PCM cooling, which uses the process of absorbing heat during the phase change to balance the heat that the battery generates, is costly but o ffers a pollution-free, high-return solution. By comparing cooling performance under di fferent conditions, Kizilel et al. [17] proved that the PCM cooling method is superior to AC in regard to economy, e ffectiveness, and safety. Through simulation and experiment, Huang et al. [18] proved that the cooling system of thermal-assisted expanded graphite is superior to the AC system in actual and extreme conditions. In addition, Wang [19] and Wang [20] greatly improved the performance of the cooling system without changing the cooling mode, by optimizing the battery pack structure. Yang et al. [21] found the specific parameters of the battery pack through optimization analysis. They pointed out that the cooling system works best when the height of the battery pack is 34 mm and the width is 32 mm.

In this study, a cylindrical BTMS based on AC is proposed. The battery pack consists of 20 cylindrical battery modules of type 18650 batteries. First, the dynamic model was built and the validity of the model was verified. Then, the BTMS of the cylindrical battery pack was optimized for the comparative analysis of di fferent battery configurations with di fferent cell layouts, di fferent positions of the air inlet and outlet in the AC system and the number of inlets and outlets. Finally, the accuracy and e ffectiveness of the whole process were verified experimentally, which provided an ideal design for the BTMS of the cylindrical battery in the future.
