1. Overview Extrusion blow molding is one of the main molding methods for producing hollow plastic parts. It is suitable for polymers such as PE, PP, PVC, thermoplastic engineering plastics, and thermoplastic elastomers, as well as various blends. It is mainly used for molding packaging containers, storage tanks, and large drums, and can also be used to mold industrial parts for the automotive industry. Like other plastic hollow molding methods, extrusion blow molding's main advantages are low production cost, simple process, and high efficiency. However, its prominent disadvantage is the difficulty in controlling the wall thickness and uniformity of the finished product. Extrusion blow molding involves placing a semi-molten plastic preform (parent) while hot into a mold of various shapes, and immediately introducing compressed air into the preform to inflate it, causing it to adhere tightly to the mold cavity wall. After cooling and demolding, a hollow part is obtained through thermoforming. The entire molding process can be divided into three stages: preform formation, preform inflation, and cooling and solidification. Researchers both domestically and internationally have been striving to study the various stages and the entire process of extrusion blow molding using different methods since the 1960s. Generally speaking, these methods can be broadly categorized into two types: experimental research and numerical analysis techniques. Numerical analysis is based on three fundamental equations: continuity equation, motion equation, and energy equation. It requires numerous assumptions to simplify these equations and is solved using finite difference or finite element methods. Furthermore, some rheological parameters in the constitutive equations are difficult to obtain. For complex-shaped work-in-process products, it requires a significant amount of computer time. Experimental research, on the other hand, is the simplest and most direct method. The following is a review and analysis of the experimental research status of each stage of extrusion blow molding. 2. Research Status of the Preform Forming Stage Preform formation refers to obtaining a semi-molten plastic tube preform (preform) through extrusion molding. As the geometry of hollow blow-molded parts becomes increasingly complex, a well-designed preform is crucial for obtaining parts with the required wall thickness distribution and structural stability with minimal material consumption. This means that in the preform forming stage, adjusting the wall thickness distribution shape of the preform is used to make the wall thickness distribution of the blow-molded product more uniform. Due to factors such as extrusion expansion, sagging, and springback during preform formation, the dimensions of the preform are inconsistent in the length direction during the preform forming stage, making it very complex. Since the temperature of the extruded polymer preform is high and cannot be directly measured, experimental research on the preform forming stage in extrusion blow molding mainly involves designing experimental methods to measure the preform diameter distribution and wall thickness distribution. The earliest experimental method to obtain the preform dimensions was used by Sheptakr et al. They designed a special mold called a "preform clamping mold" to analyze the preform. This device can only obtain the mass expansion Sw of the preform, but cannot directly obtain the diameter and wall thickness expansion of the preform. Kalyon et al. [2] added a camera device to the above device, which can be used to take pictures of the preform before the mold clamps the preform, thereby obtaining the diameter distribution of the preform. This method can obtain a more accurate preform diameter distribution, but it is time-consuming and cannot be used for online measurement, thus limiting its practical application. Another method to measure parison expansion is to directly extrude the plastic melt into oil with the same temperature and density as the melt. This allows for the measurement of parison expansion without sag or solidification. Meanwhile, since the sidewall of the oil tank is transparent glass, the parison can be photographed at certain time intervals. Due to the transparency of the plastic melt, the diameter distribution inside and outside the parison can be determined from the photograph. Due to parison expansion, the shape and size of the parison are not consistent along the parison length. To mark the location of data measurement, carbon black particles are sprayed onto the parison surface at fixed intervals using an inkjet device. However, this method does not consider the effect of sag and is difficult to apply in actual production. With the development of image analysis technology, more and more researchers prefer to use image analysis technology to determine parison size. The diameter distribution of the parison can be directly measured by images, but the thickness distribution of the parison cannot be measured directly and can only be calculated indirectly. Many researchers have tried to calculate the parison wall thickness distribution using different measurement methods and algorithms. PLSwan et al. [3] designed a device using two cameras to measure the parison expansion size (as shown in Figure 4). The parison is extruded into a container at the same temperature as the parison. The camera (9) located below is aimed at the end of the parison, while the camera (5) located above sends a signal to control the position of the camera (9) through a computer to ensure that it is always aimed at the end of the parison during the parison extrusion process. The diameter and wall thickness of the parison can be obtained from the image. However, the experimental equipment is complicated and only considers the isothermal situation, so it is not widely used in practice. RWDiraddo and A. Garcia-Rejon[4] proposed a non-contact method for measuring the parison wall thickness distribution based only on image analysis. This experiment only uses one camera to take pictures of the parison extrusion process, measures the relationship between the parison length and time, the parison diameter distribution, the extrusion flow rate, and the temperature gradient of the parison along the length direction, and then calculates the parison wall thickness distribution based on the relationship between the parison wall thickness distribution and these parameters. RWDiraddo et al. used this method to study the effects of flow rate, melt temperature, and die gap on the parison wall thickness distribution of HDPE resins with different molecular weights. This method is theoretically complicated and the experimental data processing is cumbersome. WIPatterson and MRKamal[5] developed an online closed-loop control system for the wall thickness distribution of a billet. In this system, the length and diameter of the billet can be obtained directly by a camera and an image analyzer connected to it, while the wall thickness distribution of the billet is calculated by geometric relationships, but the empirical parameters used are difficult to obtain. To achieve online closed-loop control of the wall thickness distribution of the billet, a method that can directly measure the wall thickness distribution of the billet online is needed. Assuming that the melt flow rate is constant, the wall thickness of the billet can be calculated by a simple method and can be used for online measurement. The first to use this method was Kaise in Germany, which was later improved by Svein Eggen and Arne Sommerffeldt[6]. A simplified diagram of the measuring device is shown in Figure 5. It consists of a camera, a device for spraying ink onto the surface of the billet, and an image analyzer. The diameter distribution of the billet can be obtained directly from the captured image, and then the distance between adjacent ink dots is measured. Based on the assumption that the flow rate is constant, the wall thickness distribution of the billet can be calculated. Where R is the radius of the billet, q is the flow rate, ρ is the melt density, and z is the distance between adjacent ink dots. This method is theoretically simple, the experimental setup is simple, and the measurement accuracy is high, but the experimental data is large and the processing is cumbersome. Some researchers use optical methods to study the forming of blanks. PLSwan, MR Kamal and A. Garcia-Rejon[7] developed an optical sensor measurement device, as shown in Figure 6, which can measure the thickness distribution of the blank online before the mold is closed. The device is designed based on the principle of light reflection in optics. A laser beam is directed at the surface of the blank at a certain angle. The laser beam is reflected by the inner and outer surfaces of the blank to form two laser beams. The camera lens detects this interval and sends it to the computer analysis system. According to the geometric relationship, the computer can calculate the wall thickness distribution of the blank. However, there is also the problem of light refraction at the same time as light reflection. The refraction of light cannot be ignored in this measurement method. Taking the refraction into account and determining the refractive index of the blank undoubtedly adds a lot of complexity and difficulty to this measurement method. 3. Research status of the preform blowing stage Preform blowing refers to placing the plastic tube preform in the mold while it is hot, and immediately blowing it with compressed air to form it tightly against the mold cavity wall. The forming at this stage directly affects the shape, wall thickness uniformity and performance of the product, and is the key link in the whole forming process. At this stage, the experimental research on preform blowing mainly includes two aspects: one is the study of preform blowing dynamics, and the other is the measurement of the wall thickness of the preform after the preform blowing is completed. The earliest experimental setup for studying the preform blowing dynamics was established by Musa R. Kamal, Victor Tan and Dilhan Kalyon[8]. They designed a transparent blow molding mold and used two cameras to photograph the expansion behavior of the preform in the mold. The simplified diagram of the device is shown in Figure 8. The images were sent to a graphic analyzer for analysis to determine the relationship between the diameter distribution of the preform and the time. Ryan and Dutta[9] used camera technology to monitor the free expansion behavior of the preform under moldless conditions and obtained the expansion size of the preform. Most researchers subsequently used similar methods to study the inflation behavior of preforms. Wagner and Kalyon[10] designed an internal solid pressure sensor based on Kamal[8], as shown in Figure 8. It can measure the pressure during preform inflation. At the same time, another pressure sensor is installed on the flash of the mold cavity. Thus, the two sensors can measure the real pressure difference between the inside and outside of the preform during the blowing process. They used this device to study the effect of three types of PA-6 on the inflation behavior under inflation pressure. Recently, Yong Li et al.[11] used a high-speed optical measurement system that can measure the instantaneous surface shape to measure the inflation behavior of polymer films. Its measurement diagram is shown in Figure 9. The two ends of the polymer film preform are fixed between two plates, and compressed air is introduced into the pressure chamber to make the polymer film preform inflate. The optical probe contains a CCD camera and a grating emitter. During measurement, the grating emitter emits a grating onto the surface of the polymer film preform. The grating deforms with the deformation of the polymer film preform, so the grating image contains information about the surface shape of the polymer film preform. The expansion size of the polymer film preform can be obtained by quickly capturing the grating image and sending it into the computer for processing. MCDL is a multi-channel data transceiver that can simultaneously acquire pressure and grating image signals to obtain the relationship between pressure and the shape of the polymer film preform during the expansion process. Experiments have shown that its measurement accuracy is much higher than that shown in Figure 7. There are offline and online measurements for preform wall thickness. Offline measurement is more commonly used because it is simple to measure. Offline measurement includes infrared, ultrasonic and micrometer measurements. These methods are not only time-consuming, but also require correction of deviations caused by the time lag caused by offline measurement, resulting in inaccurate measurements and many defective products. Online measurement of product wall thickness can minimize the lag time, thus improving the accuracy of correction of deviations caused by the processing process. Diderichs and Oeynhauser[12] used an ultrasonic sensor placed in the mold to measure the wall thickness distribution. Its measurement principle is shown in Figure 10. Short ultrasonic waves generated by the piezoelectric crystal in the ultrasonic sensor are reflected by the object wall and then return to the sensor. The wall thickness s of the object being measured is equal to the speed of the ultrasonic wave in the object multiplied by half the time required for the ultrasonic wave to travel in the object. However, the accuracy of ultrasonic measurement is greatly affected by polymer properties (such as density and crystallinity) and temperature. 4. Research progress on product cooling and curing stage Product cooling and curing refers to the process of cooling the preform after it is blown and pressed tightly against the mold wall by means of the mold with high thermal diffusivity and compressed air, cooling to a certain temperature, opening the mold, and then cooling in the air. It generally includes external cooling (heat conduction between the outer surface of the product and the mold cavity), internal cooling (convective heat transfer between the inner surface of the product and the cooling air or other media) and cooling after opening the mold (natural convective heat transfer between the inner and outer surfaces of the product and the air or other media). Experimental research on product cooling and curing stage mainly involves measuring the transient temperature, shrinkage rate, warpage, etc. of the product. The transient temperature of the product is generally measured by using a highly sensitive thermocouple and a data acquisition device. In 1981, Edward[13] et al. designed the "half-bottle molding experiment" to verify the theoretical prediction of its extrusion blow molding cooling process. As shown in Figure 11. In the experiment, the instantaneous temperature of the outer surface was measured by thermocouples, and the temperature of the inner surface was measured by radiation pyrometer as soon as the product left the mold. The results were basically consistent with the theoretical predictions. In 1995, Diraddo et al. [14] used six thermocouples inserted into different parts of the product at different thicknesses from different parts of the mold, and collected the temperature through the temperature acquisition device connected to them to obtain the instantaneous temperature at different thicknesses of the product. This was a significant improvement over measuring only the inner and outer surface temperatures. The earliest measurement of the shrinkage rate of the product was by Diraddo et al. [14]. They processed a grid with a size of 5mm x 5mm in the mold cavity. After the preform was blown, the grid was imprinted on the surface of the product. This allowed them to directly measure the axial and circumferential shrinkage of the product, and then calculate the radial shrinkage according to the law of conservation of mass. The warpage of the product was generally measured by a three-dimensional laser digital system [15], and then the shrinkage and warpage of the product were obtained. 5. Conclusion Experimental research has always been the most direct method to guide engineering applications, and it is also the basis and foundation of theoretical research. The extrusion blow molding process includes three stages: preform forming, preform blowing, and product cooling and solidification. Researchers in various countries are employing different experimental methods and equipment to study each stage of extrusion blow molding. This research and development is significant for optimizing processes and mold structures, as well as improving production efficiency. With advancements in technology and improvements in experimental methods, experimental research on the extrusion blow molding process will reach new heights, providing better guidance for actual production and enabling the production of hollow blow-molded parts that meet societal needs in terms of quality, performance, and other aspects.