The Heat Transfer and Fluid Flow Laboratory is capable designing cooling headers for interstand cooling at hot strip mills. The design is optimized by using a laboratory linear test bench where cooling intensities are measured. A numerical model of a finishing mill can predict the final temperature drops for various combinations of cooling headers.


Spray cooling sections are designed to reach sufficient and controllable cooling intensity. Cooling of flat products (strips, sheets and plates) without thermally inducted distortion requires high uniformity of cooling. The Inter Stand Cooling (ISC) helps to keep optimal temperature of products therefore improves the final mechanical properties of the strip and increases the productivity of the mill.

The most difficult problem when designing spray cooling sections is to keep cooling uniformity with variability of cooling intensity. It was proved that the spray cooling intensity cannot be related only to the water impingement density. The exact cooling intensity and distribution is hard to predict thus the spray cooling must be studied experimentally.

Example of water impingement distribution for interstand cooling design


The Heat Transfer and Fluid Flow Laboratory developed a test bench for simulation of cooling sections in a real plant. It allows experimental study of spray cooling for various nozzle types, pressures or header configurations for coolant temperatures in range of 20-90°C. The cooling intensity given by heat transfer coefficient distribution is gained from the experiments and used to create a numerical model of the temperature field in the cooled material. The numerical model predicts the controllability of the cooling device and assists in the preparation and adjustment of newly designed or redesigned cooling units in a real plant.

The test bench for cooling investigation allows spraying surfaces in variety of positions

Experimental procedure:

  • A steel sample is heated to an initial start temperature in an electric furnace.
  • The heated sample placed on the trolley is set up into the spraying position.
  • The pump for the water gets going and the trolley controlled by computer runs several times through the cooling section under given conditions.  
  • The temperatures recorded by thermocouples in the steel sample are saved to data logger together with the trolley position.
Scheme of the horizontal test bench, 1 – headers with nozzles, 2 – pressure gauge, 3 – test plate, 4 – motor moving trolley, 5 – girder carrying trolley, 6 – movable trolley, 7 – data logger, 8 – heater, 9 – water tank, 10 – pump, 11 – control valve
Spray cooling - laboratory measurement
Spray cooling - plant application
Various types of cooling headers can be used when optimal cooling is designed

Experimental Results

The result gained from the heat transfer test is the temperature history in the thermocouple positions. The temperatures are used to determine boundary condition, which is the heat transfer coefficient in dependency on time and position by using an inverse heat conduction problem (IHCP). The boundary condition inputs to numerical simulation of product cooling in the real plant conditions.

Heat transfer coefficient computed for each test plate pass under a nozzle configuration
The numerical simulation of the temperature and the cooling rate in dependency on the time for cooling of 3 mm steel strip under 17 rows of nozzles; simulation based on the experiment with 5 rows of flat nozzles shown in photo and boundary conditions in the right figure
A numerical simulation of interstand cooling of steel strip with final thickness of 3.15 mm at continuous finishing line; there are two examples – without cooling and with all of the interstand headers on

Leidenfrost Temperature

When the temperature drops and the cooling rate suddenly increases during spray cooling, it signifies the Leidenfrost temperature. One of the typical spray cooling record shows the Leidenfrost temperature at 800°C, the heat transfer coefficient (HTC) about 400 W/m2K for the surface temperatures above Leidenfrost temperature and about 4000 W/m2K for the surface temperatures below the Leidenfrost temperature. In such case if one surface point is at a temperature of 900°C and a neighboring point is at a temperature of 700°C, the colder point will be cooled ten times more intensively. This effect causes a significant increase of temperature non-uniformity inducted by initially small temperature differences.

Typical temperature history and cooling intensity during spray cooling test

There are only a few ways to avoid this issue. The air cooling can be used because shows no Leidenfrost phenomenon, but the high cooling rate is problematic to achieve. Another solution is to involve mist nozzles and water nozzles with fine atomizing that provide almost constant cooling intensity above the Leidenfrost temperature. However it must be taken into an account a transient area between the film boiling regime and the nucleate boiling regime. There is no possibility of a reliable numerical or empirical prediction of this area for new cooling configurations or nozzles settings. The only reliable method is the heat transfer measurement for the designed configuration and parameters.

The heat transfer coefficient of a mist nozzle as a function of surface temperature for growing water impingement densities [l/min]; the lower curve denotes the lowest water impingement density, the upper curve denotes the highest water impingement density; the danger area plotted in grey is in the transient area between the film boiling regime and the nucleate boiling regime

Impact Pressure Measurement

Impact pressure is a one of the basic nozzle spray parameters. To obtain impact pressure distribution on the sprayed surface, a test bench was developed in Heat Transfer and Fluid Flow Laboratory.

Impact pressure measurement of full cone nozzle in distance of 200 mm from the surface covered by water layer with depth of 5 mm, 20 mm and 40 mm (from left to right)

Additional Information

can be found in Publications of the Heat Transfer and Fluid Laboratory.