Measures to Reduce Corner Cracks in Ingots and Surface Defects in Rebar

1 Introduction

A steel company’s product line includes four JIS standard rebars: SD295, SD345, SD390, and SD490. Product diameters range from 10 to 41 mm, encompassing a total of 11 sizes. Due to the complex shape of rebar, surface inspection using a magnetic flux leakage detector (MFL)—a process in which the probe moves along the product surface—is challenging. Therefore, surface defects were previously primarily detected by operators through visual inspection. In recent years, the company has introduced visual inspection equipment based on image processing and surface temperature measurement technologies. Based on user needs, a device has also been added that measures the surface temperature of rolled products and identifies abnormal areas on the product surface based on temperature differences. With the implementation of these devices, all product surface inspections have become mechanized, effectively preventing the undetection of defective products. While this measure has effectively prevented the flow of defective products, the inspection process has reduced production line utilization and output. To prevent defective products from being shipped and to identify and control the causes of abnormal surface defects in rebar, the company conducted an in-depth investigation into the causes of these surface defects and implemented targeted treatments for the continuous casting defects that cause them.

2 Cause Investigation

There are many types of surface defects found on continuously cast slabs, each caused by a different cause, as shown in Table 1. The slab surface was carefully observed, and specific surface defects were thoroughly investigated.

Table 1 Types and causes of surface defects in continuous casting slabs

Approximately 99% of defects detected by product defect detection equipment exhibit characteristics similar to peeling of the surface of rebar. To determine the location and timing of these defects, a detailed investigation of the material during rolling revealed signs of peeling. Further analysis revealed that these defects corresponded to the corners of the ingot. Furthermore, signs of peeling were observed during inspection of the ingots during the initial rolling process. It is inferred that the root cause of these peeling defects lies in initial surface defects.

A detailed examination of continuously cast billets exhibiting a high number of peeling defects revealed transverse corner cracks along the bottom of the oscillation grooves at the corners of the ingots. During rolling, the rebar produced from these ingots developed peeling defects at locations corresponding to the cracks in the corners.

One of the causes of these transverse corner cracks is believed to be the formation of the ingots during the straightening process in the continuous casting straightening machine, a phenomenon that is influenced by both the composition of the molten steel and the temperature of the ingots as they pass through the straightening machine rollers. Despite adjustments to the water flow rate of the secondary cooling nozzles and the cooling position of the secondary cooling chamber, the incidence of warping and peeling defects on rebar was not effectively reduced. To precisely determine the location of transverse cracks at the corners of the billets, special measures were implemented during the continuous casting process: production was halted, the billets were suspended and stationary in the secondary cooling chamber, and sampling was performed directly in the secondary cooling chamber, bypassing the straightening rollers. The results showed that transverse cracks at the corners were already present before the billets were straightened by the straightening rollers of the straightening rollers.

The investigation to date has clearly established that warping and peeling defects on the surface of rebar are caused by transverse cracks that appear at the corners of the billets during the initial stages of continuous casting. This discovery has become a key goal for improving billet surface defects.

3 Improvement Efforts

3.1 Reducing the Impact of Vibration on Defects

Transverse cracks at the billet corners typically occur along the bottom of the vibration valley. Therefore, it is hypothesized that effectively suppressing mold vibration will significantly reduce the incidence of billet corner cracks. Kawakami et al. reported that at the end of the negative slip time (i.e., the time during which the mold’s downward vibration causes its descent velocity to exceed the billet casting speed, hereinafter referred to as tN), a vibration mark is formed by pushing and bending the tip of the solidification angle. A larger tN value results in a more pronounced pushing and bending effect and a deeper vibration mark. Conversely, a smaller tN value results in a shallower vibration mark, effectively suppressing the formation of transverse cracks in the billet. The tN value can be shortened by increasing the mold vibration frequency and reducing the amplitude.

Based on this, the mold vibration parameters were adjusted, as shown in Table 2. Maintaining a tN value of 0.1 s is considered a critical value. Therefore, considering the fluctuation of the continuous casting casting speed, the tN value is set above 0.1 s.

Table 2. Crystallizer vibration parameters

By adjusting the mold vibration parameters, the depth of vibration marks was reduced by 30%. Furthermore, after optimizing the mold vibration settings, the incidence of surface peeling defects (such as peeling) on rebar of major specifications and sizes decreased by approximately 0.6 cases per ton of rolled product. Specific data are shown in Tables 2 and 3.

Table 3 Vibration mark status before and after changing the crystallizer vibration parameters

3.2 Changing the Mold Dimple Pattern

When a large number of peeling and shedding defects occur in rebar during hot rolling, prominent points formed by dimples on the mold’s inner surface can be clearly observed on the billet surface. Based on this phenomenon, the correlation between mold dimples and corner cracks in the billet was investigated.

The factory uses a dimple-designed mold to effectively prevent billet deformation and the initiation of internal cracks. Multiple dimples are precisely machined into the inner surface of the mold, just below the meniscus.

During the early stages of solidification, these dimples create air gaps between the outer surface of the billet’s solid shell and the inner surface of the mold, inhibiting heat transfer from the solid shell to the mold. By suppressing undercooling of the initially solidifying shell, the solidification shrinkage of the solid shell within the mold is reduced, further suppressing the formation of air gaps. This ensures that the outer surface of the solidified shell conforms evenly to the four inner walls of the mold, ensuring uniform cooling on all four sides of the billet and effectively suppressing billet deformation. The closer the dimples are positioned to the meniscus, the more pronounced their cooling inhibition effect and the more distinct the raised projections transferred to the billet surface. The mold features four rows of dimples, with seven in the first row and six in the next. When raised projections appear on the billet surface, they typically appear as raised point marks left by the seven dimples. To verify the impact of these raised point marks on transverse cracks at billet corners, mold tests were conducted using different dimple treatment methods.

Also, data on the detection rate of peeling defects using a conventional mold and a mold with modified dimple parameters are provided (see Table 4).

Table 4. Operation results of changing mold pit parameters

Therefore, using a mold without dimples significantly reduces the occurrence of peeling defects, but rhombohedral deformation and related internal cracks in the billet still occur. Lowering the height of the mold’s machined dimples effectively suppresses peeling defects and also controls rhombohedral deformation in the billet.

It is speculated that when the extremely thin primary solidified shell passes through the dimples in the upper mold, thermal expansion causes traces of the dimples to form on the billet surface. During the billet drawing process, these protruding bumps become lodged in the mold’s dimples, causing cracks along the vibration marks at the billet corners.

3.3 Improving the Response Accuracy of the Mold Level Control

The implementation of the two aforementioned measures significantly reduced the incidence of transverse cracks at the billet corners, but the incidence of peeling defects during rolling increased significantly. However, significant peeling defects still occurred.

To identify the cause, a detailed review of the continuous casting records of billets with a high incidence of defects was conducted. Results showed significant fluctuations in the molten steel level within the mold during the casting process. The traces of raised points left by the pits are often clearly visible over a short length of these ingots.

In the steel mill’s continuous casting machine, thermocouples are installed on the outer wall of the mold’s meniscus to measure the position of the meniscus, or the position of the molten steel level. When the amount of molten steel being cast fluctuates, causing the molten steel level to deviate from the set position, a proportional-integral-derivative (PID) control scheme is employed to stabilize the steel level by adjusting the speed of the straightening rolls in the straightening machine.

In conventional control settings, the control coefficient is often set small to suppress sudden fluctuations in the straightening roll speed. However, this often results in the molten steel level at the meniscus taking over 10 seconds to return to the set position. Therefore, even with improvements to the mold’s pitting design, the desired effect was still not achieved due to the fluctuating steel level.

To address this issue, the PID tuning value was gradually adjusted, increasing the control coefficient to ensure that the molten steel level at the meniscus returned to the set position in the shortest possible time. By increasing the control coefficient, the casting speed fluctuation increased slightly, but the mold level fluctuation was successfully suppressed, ensuring a smooth and trouble-free continuous casting process. Subsequently, the parameters for the second test phase (Table 5) were used. By optimizing the mold pit pattern and improving the response accuracy of the mold level control, the incidence of scale lift defects was significantly reduced to approximately 0.3 cases per ton of rolled product. The investigations and countermeasures conducted in this paper revealed the following:

1) In some cases, billets cast using a pit mold still retain raised point marks from the top row of pits.

2) Billets with significant raised point marks from the pits exhibited a high incidence of scale lift defects during rolling, characterized by transverse cracks caused by vibration marks at the billet corners.

3) As the distance between the meniscus and the location of the first layer of mold pits increases, the raised marks on the billet surface gradually become thinner, and cracks at the billet corners also decrease. The range of variation for this effect is very narrow (approximately 5 mm). Based on this, the mechanism of corner cracking in a dimple mold is hypothesized to be as follows:

1) When the dimple is machined directly below the meniscus, the initially very thin solidified shell is pressed against the mold wall by the ferrostatic pressure of the molten metal, transferring the dimple mark to the surface of the billet, forming a raised point.

2) When this transferred raised point passes through the mold dimple, it catches the shell and creates resistance in the casting direction.

3) This resistance in the casting direction is converted into tensile stress, inducing cracks at the valley of the oscillation mark, which in turn forms horizontal corner cracks. This phenomenon is believed to occur during the early stages of solidification of the very thin solid shell. Stabilizing the mold level and maintaining an appropriate distance between the meniscus and the dimple machined location can also suppress the occurrence of corner cracks horizontal to the oscillation mark.

Table 5 Changes in the settings of the proportional-integral-derivative control of casting speed

4 Summary

To reduce horizontal corner crack defects in ingots and prevent peeling and peeling defects on the surface of threaded rebar, the following measures were implemented:

1) The mold vibration frequency was increased and the vibration range was reduced, namely, a high-frequency, low-amplitude mode was adopted.

2) The distance between the meniscus on the inner surface of the mold and the top row of pits was increased to 20 mm.

3) The PID settings of the straightening machine’s speed control were adjusted to improve the response accuracy of speed changes.

The results showed that the defect rate of threaded rebar of major specifications was reduced by 0.9 cases per ton of rolled product.