Antonio Augusto Gorni, Júlio Márcio Silveira e Silva, Jackson Soares de Souza Reis,
Celso Gomes Cavalcanti

Companhia Siderúrgica Paulista - COSIPA, Brazil


COSIPA steelworks, in Brazil, produces 3.9 million tons of steel per year. Its main products are plates and strips hot and cold rolled. Controlled rolled of thick gauges are performed in a single stand four-high reversing mill. The operations of roughing and finishing are executed in the same stand. This procedure limits the output of the mill considerably. There is however a fast cooling facility subsequent to rolling which can be used for improvement of productivity.

The production of microalloyed steels aims basically at reducing ferrite grain size. By doing this, it is possible to gain strength from a polygonal ferrite matrix without reducint weldability. Additionally, no off-line heat treatment is necessary in the as-rolled product. However, controlled rolling practice does introduce a decrease in tonnage off the mill, since an interval of time between roughing and finishing has to be introduced. Here, the material has to be cooled down, so that no recrystallization will occur between passes in the finishing schedule. Intermediate Forced Cooling (I.F.C.), that is, cooling between roughing and finishing, is one of the options available to increase productivity [1-7].

There is no doubt that the use of forced cooling during the holding step of controlled rolling would shorten this idle time. However, there was some concern about the influence of this new process over the final properties of the plate. An excessively severe cooling can result in a pronounced temperature gradient along plate thickness. In this case, its surface would be over-cooled during the application of the water sprays. After the cooling process, the surface would be submitted to a recalescence, as heat would flow from plate core. This thermal cycle can lead to excessive precipitation of microalloy carbonitrides, affecting significantly the ferrite start temperature and austenite recrystallization kinetics, resulting in the formation of elongated and deformed superficial grains. These alterations in the microstructure of the plate surface would not change significantly its overall properties, but could induce tensile residual stresses in its surface, which could produce surface cracks depending on their magnitude. This "chilled" superficial microstructure can also, in some cases, increase the transition temperature determined by Charpy impact tests.

The present work examines the effects of I.F.C. on mechanical properties of microalloyed steels. Firstly, experiments at laboratory conditions were carried out. These gave preliminary information on the process parameters needed to run a mill trial. Then an industrial scale experiment was carried out with a range of chemical compositions and cooling rates. No deterioration of properties was observed and a significant decrease in holding time was obtained.


Table I shows the chemical compositions of the steels used in this work. Steel 1 is the base material. It has no microalloying addition and typically shows levels of C of the magnitude of 0.14 and Mn of 0.86 (numbers in weight percent). Steels 2 to 4 are microalloyed grades. material 2 and 3 differ slightly in the Nb content. Steel 4, however, is similar to 3, but with Ti added to refine austenite grain size in the reheating furnace.

Steel C Mn Si Cu Cr Ni Nb Ti
Table I: Chemical compositions of steels studied in this work, weight percent. All steels have in average 0.040% Al, 0.018% P and 0.010% S.

Simulation of controlled rolling was carried out firstly in a laboratory rolling mill. Here steel 2 was used and simulation began by pre-heating the samples to 1070oC (approximate temperature of end of roughing at COSIPA Plate Mill) for 35 minutes. Approximately 0.013% Nb in steel 2 was solubilized after pre-heating. Specimens were then deformed in a single roughing pass of 25% reduction at 1070oC. This gave an homogeneous austenite microstructure as the one expected at the end of the industrial roughing schedule. Rolling was followed by a holding period in which six different colling conditions were applied to the sample. Specimens were then cooled, inside a furnace, in air, in water (quenching rig) or for 5, 10 and 15 s in a water spray. These conditions produced cooling rates in the range from 1.1 to 3.5o/s. Finishing was then simulated by applying three passes (31, 27 and 29% reduction) at temperatures of 915, 850 and 760 oC, respectively.

Industrial trials were performed for all four steels of Table I. The plant I.F.C. facility was used with 80 or 100% of its capacity, 167.0 l/s at 0.30 MPa (that is, 10,000 l/min at 3 kgf/cm2) for one or more cooling passes. The plates were reheated, deformed in roughing and finishing stages according to the usual controlled rolling practice at COSIPA. This procedure was employed in all plant experiments for the sake of comparing final productivity.


Laboratory scale results showed a slightly refined ferrite microstructure for the more severe cooling conditions only. The refined grains were evenly distributed accross the thickness of the sample. Formation of mixture of refined and deformed ferrite grains was not observed.

Results of tensile tests from samples hot rolled in the laboratory mill showed no deterioration of mechanical properties. On the contrary, some gain, mainly in the yield strength values, could be verified as the cooling conditions became more severe. Additionally, Charpy notch tests carried out in the laboratory samples did not show any variation in the absorbed energly values at 0oC.

Results of industrial scale experiments also showed no significant changes in mechanical properties and microstructural characteristics when I.F.C. was introduced. Cooling rates employed in the I.F.C. experiments ranged between 3.5 and 5.0oC/s. There was no significant change in the values of absorbed energy and transition temperature obtained from plates controlled rolled with or without the use of I.F.C. Also D.W.T.T. tests were carried out in the samples from I.F.C. experiments. Here, no change in toughness properties were observed. Furthermore, bending tests results indicated no detrimental effects resulted from the introduction of I.F.C. practice. Finally, the microstructure from samples with and without I.F.C. were very similar in transverse and longitudinal directions.

The I.F.C. product showed good flattening conditions. I.F.C. practice did not introduce any significant change in rolling loads values at the finishing passes. Moreover, in view of the standard operating conditions for individual plates at COSIPA, considerable reduction of processing time (40%) could be obtained. Tandem rolling practice could also be improved by introduction of I.F.C. Productivity of tandem rolling using 30 mm intermediate thickness plates could be increased by approximately 10%.


The objective of this work was to increase the productivity of the controlled rolling process, decreasing the time spent during the holding phase between the roughing and finishing steps through the use of forced cooling.

The steps of this development were:

  1. Simulation of the Process in Pilot Scale (January 1986 to October 1986)
  2. Revamping of the Cooling Spray near the Plate Mill (February 1987 to September 1988)
  3. Industrial Trials (October 1988 to January 1990)
  4. Used routinely for Shipbuilding Steels (since January 1990)
  5. Used routinely for A.P.I. Plates for Pipes (since May 1994)

This development allowed a reduction of 40% in the delay time during the holding phase of controlled rolling, and up to 17% decrease in the total time spent during the controlled rolling of shipbuilding and A.P.I. plates.


The main conclusions of this work can be summarized as follows:

  1. Introduction of I.F.C. in controlled rolling of plates yields significant reductions in processing time. There is an increase in productivity of up to 40% in individual plate rolling and of up to 10% in tandem rolling.

  2. No significant deterioration in mechanical properties and in microstructure characteristics could be observed for cooling rates of up to 5oC/s used in the holding period.

  1. MATSUDA, M. et al. Nippon Steel Technical Report, 21, 1983, 217-234.

  2. HEEDMAN, P.J. Journal of Mechanical Working Technology, 2, 1972, 117-128.

  3. HEEDMAN, P.J. & SJOSTROM, A. Scandinavian Journal of Metallurgy, 9, 1980, 21-24.

  4. FEGREDO, D.M. Metals Technology, 4, 1977, 417-424.

  5. HEEDMAN, P.J. et al. Metals Technology, 8, 1981, 352-360.

  6. ABRAMS, H. Iron & Steelmaker, 11, 1984, 11-17.

  7. SILVA, J.M.S. M.Eng. Thesis, Universidade Federal de Minas Gerais, Belo Horizonte, Brazil, 1987.

Last Update: 14 April 1997
© Antonio Augusto Gorni