Ageing Susceptibility of Continuously Annealed Low-Carbon-Steel Strips

Abstract

The continuously annealed low-carbon-steel strips from industrial production showed an average aging index of about 30 N/mm². This increase in yield strength caused by the continuous-annealing process was strongly dependent on the soaking temperature and the soaking time. Whilst the optimization of the overaging parameters and the use of a higher annealing temperature would reduce the ageing susceptibility of low-carbon-steel strips, it was considered unlikely that its magnitude could be reduced to a comparable level to that of the batch-annealed one. Within the limit of low-carbon-strip specifications, there was very little effect of the C content on the ageing. The aging index of the temper-rolled materials decreased with the increase in the temper-rolling reductions, and a maximum increase in yield strength was achieved at a temper-rolling reduction of about 1%. In this work, the effect of C content, Al content, continuous-annealing cycles, and temper-rolling reduction on the ageing susceptibility of the continuously annealed low-carbon-steel strips was examined. SEM/EBSD technique was developed to characterize the temper-rolling strain bands and to assess the effect of temper-rolling reduction on the yield strength of the low-carbon-steel strips.

1. Introduction

The development of continuous-annealing (CA) technology for steel strips was first started in 1936 [1]. However, continuous-annealing technology did not become of great interest until the first oil crisis in 1973. The basic metallurgical foundation for CA technology in low-carbon-steel strips is to use low finishing temperatures and high coiling temperatures to achieve large ferrite-grain sizes for the hot band and overaging to precipitate the carbon in solid solution [2]. The success of CA technology in strip production has been attributed not only to its uniform mechanical properties and excellent flatness but also to its very short processing time. During the short-time annealing cycle, favourable textures for deep drawing and suitable grain sizes for formability are developed by recrystallization and grain growth.

However, low-carbon-steel strips, particularly the commercial-quality grade (CQ), processed through a continuous-annealing line, as opposed to batch-annealing, are often susceptible to ageing due to the increased levels of interstitial C or N in solution. Ageing has an influence on the material shelf life, while strain ageing will lead to stretcher-strain markings, which will affect the surface appearance. The increased surface roughness would reduce the cohesion between the steel substrate and coating. Significant ageing at elevated temperature, e.g., during paint-curing heat treatment, could also lead to spring-back problems, resulting in the rejection of the products.

Ageing can be caused by the diffusion of the interstitials in solution, i.e., C and N, at elevated temperatures or even at room temperature. Except for certain stabilized interstitial-free (IF) steels, CA processed low-carbon-steel strips always exhibit a certain level of ageing. This is mainly due to the C in solid solution, since N would be combined with Al after the high-temperature coiling during hot rolling.

Temper rolling or skin passing is usually used as a means of suppressing the yield-point elongation (YPE) in low-carbon-steel strips to prevent the formation of strain bands or Lüders bands during subsequent processing. In addition, temper rolling is also used for shape correction and to impart the required material properties, such as surface roughness.

There are many studies on the ageing characteristics of CA processed deep-drawing ultra-low-carbon-steel strips [3] but very little work on commercial-quality low-carbon-steel strips. In this paper, we study the effect of four variables, i.e., C, Al, CA cycles, and temper-rolling reductions, on the ageing susceptibility of the CA commercial-quality low-carbon-steel strips.

2. Experimental Method

The low-carbon-steel sheets used in this work were CQ strips produced on an industrial CA line with a nominal soaking temperature of 750 °C for 20 s and an overaging of 350 °C for 120 s. As the materials are susceptible to natural ageing, samples received from the production line were stored in refrigeration (<−25 °C) before any further treatments or tests. The first group of A samples with variable carbon contents (from 0.015% to 0.05%) was selected with a nominal Al content of 0.039% and a standard CA cycle, while Group 2 samples contained an increased Al of 0.072% for studying the effect of Al content on ageing. In addition, Group 3 strip samples were collected from a CA cycle with a soaking temperature of 725 °C to study the effect of the CA cycle on ageing. In addition, laboratory temper rolling was carried out to achieve a range of temper-rolling reductions (0–4%).

All strip samples used in this work were cold rolled to a final thickness of 0.8 mm from hot bands with a nominal coiling temperature of 650 °C.

Table 1. Nominal compositions of the CQ strip samples used in this work (all in weight percentage).

Sample	C (wt.%)	Mn (wt.%)	N (wt.%)	Al (wt.%)	P (wt.%)	Variables
Group 1	0.016–0.05	0.180	0.0021	0.039	0.012	C contents
Group 2	0.040	0.190	0.0022	0.072	0.012	Al contents
Group 3	0.035	0.180	0.0016	0.034	0.007	Soaking temperatures
Group 4	0.030	0.172	0.0030	0.030	0.018	Temper-rolling reductions

lists all the samples with different variables.

Samples of 350 mm × 90 mm were heat-treated at a laboratory CA simulator to simulate double-annealing paint-curing cycles at 240 °C soaking for 30 s or at 225 °C soaking for 5 s (Figure 1). The temperatures were controlled by thermocouples welded on the sample, while a constant cooling rate of about 20 °C/s was achieved using compressed nitrogen.

The tensile properties were measured from at least four 50 mm gauge-length tensile specimens, with a crosshead speed of 2.5 mm/min through the yield point and 50 mm/min for the remainder of the test. The tensile tests showed good repeatability with overall experimental error within 10%. When a sample showed no obvious yielding, the 0.2% proof strength (R_p) was quoted, while, for the others, the lower yield strength (R_eL) was used. All tensile tests were conducted along the rolling direction on a Zwick 1474 universal testing machine.

Aging index (AI) is used to study the ageing susceptibility of the low-carbon-steel strips. As shown in Figure 2, AI is the difference between the flow stress at a specific pre-strain and the lower yield stress after 100 °C/1 h artificial ageing in oil bath. The increase in yield stress after the pre-straining and artificial ageing is the total hardening due to both ageing and the work hardening. The purpose of the pre-strain is to remove the yield-point elongation; therefore, the amount of pre-strain needs to exceed the yield-point elongation, varying between 2~10%. Throughout this work, a 2% pre-strain was used for measuring the aging index.

TEM carbon extraction replicas, prepared from the middle thickness of the strip samples, were examined on Philips EM400T operated at 100 kV and the associated ISIS300EDA analytical facility to study the distribution and the details of the carbide phases in the CA samples. The EBSD measurements were carried out on the cross sections of the samples along the rolling direction on a JOEL 9500 SEM with HKL system. The EBSD samples were prepared by electropolishing and scanned for an area of approximately 0.9 mm × 0.5 mm at the middle thickness of the samples using a step size of 1 µm.

3. Results

3.1. Effect of Carbon Content

Carbon is generally considered to be the main cause of ageing in low-carbon-steel strips, as nitrogen would be combined with alloying elements, such as Al or B. Samples with various C contents (0.016–0.05%) within the range of low-carbon CQ strip products were heat-treated with the two paint-curing cycles. Tensile tests were carried out on both the as-received and the heat-treated samples. The as-received CA samples showed less than 1% YPE. The average aging index between the samples with different C contents was very similar, about 32 N/mm², though the CQ strips with lower C contents showed higher total hardening than those with C contents at 0.037 or above (

Table 2. Average tensile test results on samples with various C contents.

C (wt.%)	Conditions	ReL (N/mm²)	YPE (%)	UTS (N/mm²)	A50 (%)	Ageing/AI (N/mm²)	Total Hardening (N/mm²)
0.016	As-received	226	0.2	346	36.3	AI 34	60
	Aged at 240 °C/30 s	288	4.5	341	36.8	62	-
	Aged at 225 °C/5 s	269	3.3	346	35.7	43	-
0.024	As-received	215	0.3	335	36.3	AI 33	60
	Aged at 240 °C/30 s	280	4.9	331	37.1	65	-
	Aged at 225 °C/5 s	249	3.1	335	37.5	34	-
0.037	As-received	243	0.9	349	38.1	AI 32	43
	Aged at 240 °C/30 s	303	4.9	348	34.7	60	-
	Aged at 225 °C/5 s	278	4.2	346	37.5	35	-
0.042	As-received	242	0.6	347	36.7	AI 31	45
	Aged at 240 °C/30 s	299	4.9	345	35.6	57	-
	Aged at 225 °C/5 s	282	3.8	352	34.1	40	-

).

Paint-curing treatment at 240 °C for 30 s resulted in an increase of about 60 N/mm² in yield strength and an increase of about 4.3% in yield-point elongation. The effect of C contents on the increase in yield strength and YPE was within test errors.

Samples with four different C contents were also heat-treated using a shorter annealing cycle of 225 °C/5 s. The increase in yield strength after the heat treatment was only about 34–43 N/mm². By lowering the paint-curing temperature with a shorter soaking time, the average ageing obtained from the CA low-carbon-steel strips was reduced by 30~45% (Figure 3). Though the amount of ageing appeared to decrease slightly with the increase in C contents within the range of this study, the variations were within the range of test errors, and the lines in Figure 3 were for guiding reference only. The total C content in continuously annealed CQ strips had relatively small impact on the level of ageing after paint-curing cycles, whereas the paint-curing cycles had a significant effect on age hardening of the CA low-carbon-steel strips.

3.2. Effect of Aluminium Content

The contribution of solute nitrogen on ageing of low-carbon-steel strips is about twice that of carbon in solution [4]. Therefore, it is important to remove all free nitrogen from solution in low-carbon-steel strips. Al-killed low-carbon-steel strips for CA are usually coiled at a relatively high temperature of about 650 °C to promote the formation of coarse AlN and carbides because fine AlN precipitates will pin grain growth and affect the development of favourable texture for deep drawing during continuous annealing.

Low-carbon CQ strips generally contain about 0.032% Al for deoxidization and the removal of solute nitrogen. A CA low-carbon strip with double the amount of nominal Al content was selected for tensile tests to study its ageing response. The as-received material showed a small YPE of less than 0.5%, likely due to room-temperature ageing. A simulated paint-curing treatment was carried out at 225 °C for 5 s, and the strip samples exhibited an average increase in yield strength of 28 N/mm² and an increase in YPE of about 2.7%. As shown in Figure 4, the continuously annealed CQ strips with 0.072% Al showed a smaller ageing effect than that of the strip samples with standard Al content. The tensile tests also revealed that the as-received samples with the increased Al content had lower yield strength but similar ultimate tensile strength (UTS) compared to the standard CQ strips.

The increased Al content appeared to reduce the ageing tendency of the continuously annealed CQ strips. Since the low-carbon CQ strips contained no boron, the solute nitrogen could only be removed by Al. Although the standard CQ chemistry contains a sufficient amount of Al for deoxidation and stabilizing solute nitrogen, the kinetics of forming AlN during hot-band cooling may prevent the complete removal of nitrogen. As the Al content in the strips is further increased, it is expected that the driving force for forming AlN during hot-band coiling is also increased. In fact, previous work [3] indicated that an increase in Al content in non-B-bearing low-carbon-steel strips to more than 0.05% is likely to fix more than 70% of the total nitrogen as AlN even if the hot bands are coiled at 650 °C.

3.3. Effect of CA Cycle

A typical CA cycle for low-carbon-steel strips consists of a short soaking treatment at >700 °C, followed by rapid cooling, and then overaging at about 300 °C. During the soaking stage, recrystallization and grain growth occur, and some of the iron carbides dissolve. Fine iron carbides start to form during cooling and overaging, and their density depends on both the number of nucleation sites and the degree of carbon supersaturation obtained through the rapid cooling. A CA cycle for low-carbon-steel strips is usually designed in such way to promote the formation of fine carbides and to minimize the carbon in solution.

Low-carbon CQ strip samples from two CA cycles with different soaking temperatures (i.e., Cycle 1: 750 °C/20 s and Cycle 2: 725 °C/20 s) were selected to study their ageing response. Tensile tests of the as-received materials indicated that the strip processed with Cycle 1 had no yielding, while the CQ strip processed with Cycle 2 exhibited a room-temperature ageing of 20 N/mm² after 15 days compared to the test results of the CA production line. After a simulated paint-curing treatment at 240 °C for 30 s, the samples processed with CA Cycle 1 and Cycle 2 showed increases in yield strength of 63 N/mm² and 72 N/mm², respectively (Figure 5), while both samples had similar increases in YPE of about 4.6%. Nevertheless, by pre-straining the as-received material to 2%, followed by the ageing treatment, samples obtained from the two different CA cycles showed almost identical AI values of 30 N/mm² for Cycle 1 and 31 N/mm² for Cycle 2.

Although the sample with lower CA soaking temperature showed increased yield and tensile strength, tests indicated that the aging indices of the strip samples remained the same with the two CA soaking temperatures. This might be due to the relatively small differences between the two soaking temperatures. In fact, earlier research indicated that the level of solute carbon in CA processed low-carbon-steel strips was more sensitive to the cooling rate after soaking and the overaging treatment than the soaking temperature [5]. However, a high soaking temperature will increase the level of C supersaturation during fast cooling, which will promote the precipitation of fine iron carbides and reduce the distance of carbon diffusion.

3.4. Effect of Temper-Rolling Reduction

Temper rolling can effectively remove YPE as well as control the flatness of the CA processed low-carbon-steel strips. Temper rolling does not change the intrinsic properties of the material but simply alters the response to further deformation. In this work, laboratory temper-rolling reductions of 0.5–4% have been carried out on a non-temper-rolled CA low-carbon CQ strip. The temper-rolled samples were then subjected to the simulated paint-curing heat treatment at 240 °C for 30 s.

As shown in Figure 6a, the yield strength of the as-temper-rolled materials decreased first, then increased with the increase in temper-rolling reduction. This was due to the dislocations introduced by the initial small amount of temper rolling mainly pinning the interstitial atoms. Once the temper-rolling reduction was sufficient to remove the YPE, further strain would result in work hardening, leading to the increase in yield (or proof) strength. For the low-carbon strip samples used here, a temper-rolling reduction of >1% was sufficient to remove the YPE (Figure 6b).

After a simulated paint-curing heat treatment at 240 °C/30 s, the temper-rolled samples showed a gradual increase in yield strength and a decrease in YPE with the increase in the temper-rolling reduction. However, the ageing hardening after the simulated paint-curing heat treatment increased initially with the increase in temper-rolling reduction, reaching a maximum at about 1% temper rolling. Further increase in the temper-rolling reductions led to the decrease in ageing (Figure 7). The aging-index measurements of the as-temper-rolled samples showed that the AI values remained unchanged at about 30 N/mm² and decreased to about 20 N/mm² only when the temper-rolling reductions were about 3%. Since the 2% pre-strain used for the aging-index measurements did not remove the yield-point elongation of the non-temper-rolled sample, no aging index data were obtained for the non-temper-rolled material.

4. Discussion

Processes prior to continuous annealing, such as hot-band-coiling temperature, have a significant influence on the properties of CA low-carbon-steel strips. The presence of fine AlN particles in Al-killed low-carbon-steel strips will adversely affect the development of favourable texture for deep drawing during CA. Therefore, the hot band of CA low-carbon-steel strips will be coiled at relatively high temperatures (~710 °C) to achieve large ferrite grains with coarse carbides and AlN.

The coarse-carbide phases in CQ strips were partially dissolved during the high-temperature soaking of a CA cycle, and the solute C was subsequently quenched in solid solution. The supersaturated C then precipitated as iron carbide particles during overaging. TEM examination was carried out using carbon extraction replicas, and the phase identification was primarily based on EDX analysis. The CA samples contained many small iron carbide particles within the grains and many large carbides mainly at grain boundaries, as shown in Figure 8. The large grain-boundary carbides, often co-existing with MnS phases, are believed to be the undissolved carbide phases from the hot band. In addition, small AlN particles were also observed.

The complete reprecipitation of carbon is more difficult in CA low-carbon-steel strips due to the short treatment time. As a result, the solute C and sometimes nitrogen will result in ageing. Nitrogen is generally combined with Al to form coarse AlN at hot-band coiling, and sometimes a small amount of boron (~30 ppm) will be added in low-carbon-steel strips to ensure the complete removal of solute nitrogen. Therefore, solute C in low-carbon CQ strips is the dominant factor in ageing behaviour, particularly at temperatures above 150 °C when the diffusion rate of carbon is significantly increased. In this work, the simulated paint-curing double-annealing cycles were used rather than the typical automotive paint-baking cycle (170 °C for 20 min). The relatively high annealing temperatures caused a significant amount of ageing, likely due to the further precipitation (such as ε-carbides) on dislocations and exiting particles [6]. A higher paint-curing annealing temperature with a longer soaking time (i.e., 240 °C/30 s) generated more ageing than a gentle paint-curing cycle at 225 °C for 5 s.

This study on strip samples with a range of C contents showed that the average aging index of the as-received CA low-carbon CQ strips was about 30 N/mm², with YPE varying from 0.2% to 0.9%, where the variation in YPE was likely due to the variation in temper-rolling reduction. For continuously annealed deep-drawing commercial-quality low-carbon-steel strips, a YPE of 0.2% or less is generally accepted for non-ageing products, which do not develop stretcher-strains when press formed [7]. Test results indicate that variation in the total C content in the CQ strips has very little impact on aging index, which will be predominantly determined by the level of solute carbon. However, increasing Al content to 0.072% slightly reduced the ageing tendency of the CA low-carbon-steel strips, which might suggest that the increased Al content in the low-carbon-steel strips increased the tendency of forming AlN, particularly in the hot band where the coiling temperature was at or below 650 °C [3]. Figure 9 shows the linear correlation of YPE with ageing after pain-curing treatments.

The percentage of temper-rolling reduction will determine the amount of strain in the strip, which is important for ageing or strain ageing. Temper rolling introduces a pattern of localized residual stresses. When the residual stresses reach certain levels, the yielding behaviour becomes continuous so that no stretcher-strains are developed during subsequent processing. It is estimated that a 1% temper-rolling reduction, which is typical for eliminating the yield point in low-carbon-steel strips, will plastically deform approximately half of the material [3]. Temper-rolling reduction also has direct influence on the yield strength of the low-carbon strip materials, and an optimum level of temper-rolling reduction will result in a minimum increase in the yield strength.

In this work, the optimum level of temper-rolling reduction for the CA low-carbon CQ strips was about 1–1.5%. Higher temper-rolling reduction than the optimum value would lead to work hardening but reduce the strain-ageing effect due to the increased dislocation density, resulting in a large incidence of dislocation entanglements and reducing the efficiency of carbon pinning [8].

The grain orientation spread of 0.5° was found to be the suitable value to differentiate the localized strain bands from the surrounding matrix, as shown in Figure 10. The strain bands are approximately 30~40° of the rolling direction, and the cross deformation bands increase with the increase in temper-rolling reduction.

Using EBSD grain orientation spread, we can quantitatively calculate the percentage of areas covered by the localized strain bands. As shown in Figure 11, the grain area of the strain bands increases initially with the increase in the temper-rolling reduction and then levels off to approximately 43% after about a 1% temper-rolling reduction, which is in agreement with the conclusion that a 1% temper-rolling reduction will have plastically deformed about half of the volume of the material [9].

5. Conclusions

The continuously annealed low-carbon CQ strips showed an average aging index of about 30 N/mm², which is regarded as acceptable in terms of room-temperature ageing. The increase in yield strength after laboratory-simulated paint-curing treatment was due to the return of the YPE and real ageing. In this work, the effect of C content, Al content, continuous-annealing soaking temperature, and temper-rolling reduction on the ageing susceptibility of the CA low-carbon-steel strips was examined. It is found that the ageing hardening of the CA low-carbon-steel strips increases linearly with the increase in YPE after paint-curing treatments.

Within the low-carbon CQ strip specification, the total C content has little effect on ageing, as the ageing response is primarily governed by the C in solution. Similarly, the soaking temperatures used for the two CA cycles appear to have little effect on the aging index of the low-carbon CQ strips because the solute C is more sensitive to the cooling rate after soaking and the overaging treatment. On the other hand, doubling the Al content to 0.072% slightly reduced the ageing response of the low-carbon-steel strips, which is expected to be more pronounced for CA low-carbon-steel strips with low hot-band-coiling temperature.

The yield strength of the as-temper-rolled CA low-carbon-steel strips decreased first, then started to increase with the increase in temper-rolling reduction, whereas the aging index remained unchanged until the temper-rolling reduction was above 2%. The YPE was completely removed at 1% or higher temper-rolling reduction. EBSD grain-orientation-spread mapping was used to analyse the localized strain bands induced by temper rolling. The grain area of the strain band increased with the increase in the temper-rolling reductions and reached about 43% at 1% temper-rolling reduction.

Even if the ageing response of the low-carbon CQ strips is substantially reduced by temper rolling, a return of YPE will still be observed after paint-curing treatments, and the increase in yield strength will at least reach the level exhibited before the temper rolling. The relatively small grain size of CA low-carbon-steel strips will also result in a larger yield-point elongation in the as-annealed condition and hence a larger increase in yield strength on the return of the yield point through paint-curing treatment. Whilst optimizing the CA overaging parameters and using higher annealing temperature to achieve coarse-grain size that may reduce the ageing susceptibility of the low-carbon CQ strips, it is unlikely that its ageing can be reduced to a comparable level to that in batch-annealed low-carbon-steel strips.