Transient Bottom Jet Impingement Cooling of Steel

Abstract

Accelerated run-out table cooling has become a key technology that determines microstructure and resulting mechanical properties of thermo-mechanically controlled processed (TMCP) steels. The present study quantifies the heat transfer mechanisms during bottom jet cooling of a stationary steel plate with systematic pilot-scale experiments. The emphasis of the study is to quantify the effect of process parameters, i.e. jet impingement velocity, water temperature and nozzle orientation, on heat extraction rates. Experimental results are described and quantitatively analyzed, adding to the database for run-out table cooling.

1. Introduction

The demand for high-quality steel with high strength, good weldability, durability and versatility of applications has seen a surge in the production of hot rolled thermo-mechanically controlled processed (TMCP) steel strip(s) and plate(s) in the past few decades.^1,2,3,4) The thermo-mechanically controlled process produces fine grained ferritic and/or bainitic steels with improved properties by employing accelerated cooling on the run-out table of a hot mill. The final microstructure and resulting mechanical properties are tailored by controlling austenite decomposition with suitable combinations of cooling paths and coiling or cooling stop temperatures. Hence, the quantification of heat transfer during accelerated cooling on a run-out table is pivotal for optimizing process control and productivity.

An industrial run-out table uses arrays of top and bottom water jets, and cooling is achieved by jet impingement (forced flow) boiling of water.⁵⁾ Different boiling mechanisms, i.e. nucleate, transition and film boiling, are relevant for run-out table cooling and had been established with a study on conventional pool boiling by Nukiyama.⁶⁾ A boiling curve, which is a representation of heat fluxes as a function of surface temperature characterizes these different mechanisms. Nucleate boiling is the most efficient mode of heat transfer which occurs at lower surface temperatures. Film boiling takes place at higher temperatures and is the least efficient heat transfer mechanism as an insulating vapor layer forms between the solid surface and the liquid coolant and limits the heat transfer rate. Transition boiling is a meta-stable bridge between nucleate and film boiling mechanisms. The change from film to transition boiling with decreasing temperature is associated by a minimum in the boiling curve known as Leidenfrost point whereas the change from transition to nucleate boiling occurs at the maximum in the boiling curve, known as the Critical Heat Flux (CHF).

To study jet impingement boiling, laboratory scale experiments have been conducted in the past. Experimental results show the possible existence of all the boiling mechanisms i.e. nucleate, transition and film boiling for the range of temperatures relevant to run-out table cooling, as reviewed by Wolf et al.⁵⁾ and Guedia et al.⁷⁾

Water jet impingement cooling experiments can be conducted either under steady state or transient conditions. Run-out table cooling can be simulated by pilot scale transient experiments with stationary and moving plates, respectively. Most of the transient experiments in the past have been conducted for top cooling using planar and circular jets.^8,9,10,11) Transient bottom cooling tests were carried out by Mozumder et al.¹²⁾ and Chester et al.¹³⁾ using circular jets. Morisawa et al.¹⁴⁾ did a comparative study on top and bottom cooling of a moving plate by a single circular nozzle. Top and bottom jets are hydro-dynamically different,^5,14) and bottom jet experiments using planar nozzles have yet to be explored.

Cooling in jet impingement boiling is characterized by two distinct zones, i.e. an impingement zone close to the water jet and a parallel flow zone at a distance farther from the jet.^7,15) Top cooling experiments were conducted under transient conditions by Ishigai et al.⁸⁾ on stainless steel to study heat transfer in the stagnation line, i.e. the jet centerline. Samples were cooled using a top planar jet with water temperatures between 45 and 95°C. Film boiling was observed at high surface temperatures, and heat fluxes increased with increasing jet velocity. Film boiling ceased to exist when the water temperature was lowered to 45°C. Ochi et al.⁹⁾ also observed that using water less than 35°C prevents the formation of a vapor film during transient top cooling by a circular jet for surface temperatures as high as 1000°C. Steady state experiments using a top planar nozzle conducted by Robidou et al.^16,17) showed the influence of increasing distance (up to x/w_n = 55, where x is distance and w_n is nozzle width) from the stagnation line on heat flux values, i.e. heat flux values in transition and film boiling regions decrease with increasing distance. Similarly, Nobari et al.¹¹⁾ also show the dependence of transition and film boiling with distance from jet impingement in transient cooling experiments on low carbon steel using top circular and planar jets. The maximum heat flux values drop sharply with increasing distance, before becoming nearly independent of distance, a trend also observed by Robidou et al.¹⁶⁾ Earlier, Hall et al.¹⁸⁾ had examined the transient maximum heat flux values in the boiling curves at radial positions for a top circular nozzle, and observed a sharp drop of the heat flux values with increasing distance, clearly demarcating the impingement zone and the parallel flow zone. Further, the results of Nobari et al.¹¹⁾ showed that transition and film boiling heat fluxes increase with increasing jet impingement velocities and decreasing water temperatures in the impingement zone, whereas they become independent of these parameters in the parallel flow zone.

Further, a trend called a “shoulder” in the transition boiling mechanism was first observed by Ishigai et al.⁸⁾ The “shoulder” is characterized by a region of nearly constant heat flux values over a range of surface temperatures. Ishigai et al.⁸⁾ observed this shoulder for all experiments with a water temperature lower than 75°C. The heat fluxes and width of the “shoulder” increase with decreasing water temperatures. Later, Robidou et al.¹⁶⁾ reported a shoulder up to a distance of x/w_n = 6 from the stagnation line for steady state experiments with a planar nozzle.

Transient bottom cooling pilot-scale experiments were conducted by Chester et al.¹³⁾ to study the effect of changing the nozzle inclination angle on heat extraction using a circular nozzle. An asymmetry in the flow progression was observed with distance from the center of the jet. Water progressed faster in the direction in which the nozzle was tilted with respect to the vertical axis, and slower in the opposite direction. The calculated cumulative heat extracted during a given cooling period and overall heat extracted decreased with increasing the nozzle inclination (0–30°). Wang et al.¹⁹⁾ studied the effect of nozzle inclination (0–45°) during stationary top cooling by a planar nozzle. No prominent effect was seen on the heat flux values in the impingement zone for different nozzle inclinations. However, an asymmetry in heat flux distribution was seen in the parallel flow region during cooling. Heat flux values were reported to increase along the direction of inclination with increasing nozzle inclination, and decrease in the opposite direction due to asymmetry in the flow of water.

The results from the above top and bottom cooling experiments form a database to develop models as predictive tools for optimization of industrial run-out table cooling processes.^11,20,21) Although run-out table cooling involves moving steel plates or strips, stationary plate tests provide an important benchmark for heat transfer modeling, i.e. in the limit of zero strip/plate velocity. The goal of the present study is to enhance the experimental data set for transient bottom jet cooling with systematic tests on a pilot-scale run-out table employing a planar nozzle. The present work is complementary to the transient top cooling experiments conducted by Nobari et al.¹¹⁾ on stationary plates. The effect of process parameters i.e. jet impingement velocity, water temperature and nozzle inclination, on the heat extraction rates have been quantified for the impingement zone and the parallel flow zone. Based on the experimental observations, boiling curves have been proposed for both cooling zones.

2. Methodology

2.1. Pilot-scale Facility

Transient experiments were conducted on a 15 m long run-out table facility (Fig. 1(a)) capable of simulating industrial conditions for both stationary and moving plates. A 20 kW furnace at one end of the table heats samples up to 1000°C. The maximum sample size is a length of 1.2 m and a width of 0.43 m. The furnace is fitted with a 0.48 × 1.2 × 0.13 m pocket within which nitrogen gas is supplied to form an inert atmosphere thereby minimizing scale formation during heating. After heating to a desired temperature, samples are moved under a 6.5 m tall cooling tower by a hydraulic chain pulley drive system. The cooling tower consists of an overhead tank equipped with a 30 kW immersion heater for conditioning water to a pre-set temperature value. Cooling water of the desired temperature (with an accuracy of ± 0.5°C) is circulated in a 1.5 m³ closed loop in the containment tank and can be pumped through either top or bottom header(s) up to a flow rate of 500 l/min (correct up to ± 0.5 l/min). The water temperature and flow rates are measured by a temperature probe positioned in the flowing jet and Omega FTB-905 flow meters positioned in the header inlet, respectively. Different top and bottom header arrangements with a single planar nozzle or an array of multiple circular nozzles can be installed in the cooling tower. For this study, a single planar nozzle of cross section 250 × 4 mm was installed in the bottom header (Fig. 1(b)).

Fig. 1.

(a) Schematic of pilot scale run-out table with bottom planar nozzle (b) nozzle dimensions. (Online version in color.)

2.2. Sample Preparation

As hot rolled microalloyed low carbon steel plates of dimension 600 × 430 × 6.6 mm were used for experimentation. A fresh plate was employed in each experiment to ensure uniform surface conditions. Prior to a test, the surface roughness (ISO 1997) was measured using a Mitutoyo Surface Roughness Measurement Surftest (SJ-310). The average arithmetic mean roughness (Ra) recorded on evaluation lengths of 1 cm and measured within an accuracy of ± 0.01 μm was found to be in the range of 1.5–2 μm for all plates. Flat bottom holes of 1.59 mm diameter were drilled at selected spatial locations in each plate for connecting K-type thermocouples at approximately 1 mm from the cooled surface. The depth of each hole was measured using a micrometer with a precision of ± 0.01 mm. Each thermocouple was spot welded in an intrinsic junction at the base of the hole, and a thermocouple wire pair was insulated from contact using ceramic tubes (Fig. 2). The thermocouples recorded the transient sub-surface temperatures at eight horizontal distances (Figs. 3(a) and 3(b)) from the jet centerline (x = 0 mm) during cooling with a reported accuracy of ± 2°C for T < 277°C and ± 0.75% for T ≥ 277°C.²²⁾ It is often discussed that sub-surface thermocouples record slightly higher temperatures because of the existence of an isolation material behind the hot junction. However, due to the sufficiently high thermal conductivity of steel this heat dissipates quickly around the hole, as suggested also by Franco et al.²³⁾ Hence, any possible discrepancy in recorded temperature values due to reduced effusivity in the thermocouple hole is assumed to be significantly lower than the intrinsic error of temperature measurement using K-type thermocouples. Thermocouple spacing was denser near the stagnation line where higher thermal gradients are expected, and a number of thermocouples were repeated at symmetric locations on either side of the plate centerline (x = 0 mm) for backup data.

Fig. 2.

Schematic of thermocouple spot-welded at a location on a test plate.

Fig. 3.

Schematic showing thermocouple locations with respect to jet centerline for (a) non-inclined jet (b) inclined jet. (Online version in color.)

2.3. Experimental Procedure

Based on the top cooling study of Nobari et al.,¹¹⁾ a test matrix was selected to evaluate the effect of process parameters, i.e. jet impingement velocity, water temperature and nozzle inclination, on heat extraction rates, as detailed in Table 1. The ranges of jet impingement velocity and water temperature employed in this study are equivalent to the tests conducted by Nobari et al.¹¹⁾ The jet impingement velocity (v_i) is quantified by the flow rate (F) measured during the experiment with the following equations:   

$$v_n=(F/A)/6*{10}^4$$(1)

  

$$v_i={(v_n{}^2-2gh)}^{1/2}$$(2)

where v_n is the velocity of water at nozzle exit in m/s, measured flow rate F is in l/min, A is the cross-sectional area of the nozzle in m², h is the standoff distance in m and g is acceleration due to gravity in m/s².

Table 1. Experimental test matrix.

Test setup	Impingement velocity, vi (m/s)	Flow rate, F (l/min)	Water temperature (°C)	Jet Inclination (°)
1	2.3	160	40	0
2	3.1	200	40	0
3	4.8	300	40	0
4	2.3	160	25	0
5	3.1	200	25	0
6	4.8	300	25	0
7	2.3	160	10	0
8	2.3	160	25	10
9	2.3	160	25	20

Prior to a test, the plate was heated to 800°C in the furnace. Before heating, the center of the cold plate was aligned manually with the impinging jet for each test. The nozzle to plate standoff distance was set to 88 mm for all tests. Upon reaching the set furnace temperature, the plate was moved to the cooling tower. Water in the nozzle was flowing at the pre-set flow rate and kept by a diverter from coming in contact with the hot plate prior to cooling. The plate was then cooled by the jet from an average start temperature of 700°C to about 100°C. The transient sub-surface temperature time histories were recorded at a frequency of 52 Hz by an Iotech DaqBook 2005 processor unit and transferred to a computer by DASYLab 8.0 data acquisition software.

2.4. Data Processing

The raw sub-surface temperature histories were filtered by a two-step approach proposed by Caron.²⁴⁾ First, a moving median was used to remove short term fluctuations while preserving the trend of the temperature-time curves. Thereafter, any arbitrary edges which can occur due to the use of the moving median were further filtered by a moving mean. The filtered raw data provided the input for an Inverse Heat Conduction (IHC) algorithm developed by Zhang²⁵⁾ in order to solve the second order heat transfer equation given by:   

$$\frac{1}{r}\frac{\partial }{\partial r}\left(rk\frac{\partial T}{\partial r}\right)+\frac{\partial }{\partial z}\left(k\frac{\partial T}{\partial z}\right)= \rho C_p\frac{dT}{dt}$$(3)

where ρ is the density of the material (kg/m³), k is the conductivity (W/m.°C), C_p is the specific heat capacity (J/kg.°C), T is temperature (°C), t is time (s), r and z are cylindrical coordinates. The IHC calculations were performed in a 2D axisymmetric Finite Element domain (Fig. 4) with a finer mesh near the quenched surface where higher heat flux gradients are expected. The mesh density is detailed in Table 2. The filtered raw data served as boundary condition at the measured depth from the quenched bottom surface to calculate the surface heat fluxes and surface temperatures. Heat loss by free convention and radiation served as the boundary condition on the top surface. Adiabatic boundary conditions were assumed for all other surfaces of the calculation domain. Temperature dependent thermo-physical properties of the microalloyed low carbon steel plate were employed for the FEM model.

Fig. 4.

Schematic of 2D axisymmetric finite element domain.

Table 2. Meshing details of 2D axisymmetric domain.

Section	Number of elements	Arrangement (r × z)
A	25	5 × 5
B	50	10 × 5
C	100	10 × 10

The IHC algorithm uses a function specification method combined with a zeroth-order Tikhonov regularization method, employing a sequential in-time concept utilizing the information from three future time steps for enhanced accuracy.²⁵⁾ The IHC model does not take into account the latent heat of phase transformation. However, for the range of cooling temperatures in this study it is not significant as it can be assumed that the austenite-ferrite transformation in the employed low carbon steel occurs prior to the start of jet impingement cooling. Errors in the experimental parameters lead to an uncertainty band in the output of the IHC model. Vakili²⁶⁾ had estimated uncertainties of the calculated heat fluxes with a “computerized uncertainty analysis” to be ±16% in the impingement zone and ±8% in the parallel flow zone for experimental conditions comparable to this work.

3. Results and Discussions

3.1 Boiling Curves

The IHC algorithm quantifies the surface temperature and heat flux histories at each thermocouple location. As a representative example, Figs. 5(a) and 5(b) show the surface temperature and heat flux histories of test 3 (Table 1) at four different locations with respect to the jet centerline (x = 0 mm). The time scales in Fig. 4 are taken such that t = 0 s corresponds to 1 s prior to transition from air to jet impingement cooling. During the period of air cooling, the cooling curves (t < 1 s) coincide, showing homogeneous temperature distribution across the plate. As the planar jet hits the plate, the cooling curves in the impingement zone (x ≤ 20 mm) close to the jet centerline show a sharp drop in temperatures (Fig. 5(a)) that corresponds to a sharp increase in the surface heat fluxes (Fig. 5(b)).

Fig. 5.

(a) Surface temperatures vs time (b) heat fluxes vs time of selected TC locations for test 3; impingement velocity: 4.8 m/s, water temperature = 40°C. (Online version in color.)

In the parallel flow zone at farther distances (x = 160 mm) from the jet centerline, the cooling curve (Fig. 5(a)) shows a considerable period of air cooling. The eventual drop of temperature in the cooling curve coincides with the progression of the wetting front on the surface. The heat flux in the parallel flow zone shows a peak value lower than that in the impingement zone (Fig. 5(b)).

The cooling curve at a distance between the impingement zone and the parallel flow zone (x = 40 mm) is characterized by significant fluctuations for a period of time after the jet impinges the plate (Fig. 5(a)), before an eventual sharp drop in temperature values. These initial fluctuations are also reflected in the corresponding heat flux values, before reaching the peak heat flux (Fig. 5(b)).

Plotting the transient surface heat fluxes with respect to the surface temperature for different distances depicts the family of boiling curves of test 3 (Fig. 6). Different heat transfer mechanisms during jet impingement cooling, i.e. nucleate boiling and transition boiling, can be identified. For the boiling curves in the impingement zone near the jet centerline (x = 0, 20 mm), an initial cooling region characteristic of transient experiments²⁷⁾ is seen for higher surface temperatures and coincides with the transition from air cooling to jet impingement cooling. A ‘shoulder’ in the boiling curve with fluctuating heat fluxes is observed as surface temperatures drop below 600°C. This ‘shoulder’ in the boiling curves of the impingement zone is present in all tests and is consistent with the top cooling results of Robidou et al.^16,17) and Ishigai et al.⁸⁾ A transition boiling region follows the termination of the ‘shoulder’ at lower surface temperatures until a maximum in the boiling curves is reached. For temperatures below 200°C, the heat transfer mechanism changes to nucleate boiling. The heat flux values in the shoulder and subsequent transition boiling regions drop to lower values with increasing distance in the impingement zone, whereas they appear to be independent of distance in the initial cooling and nucleate boiling regions.

Fig. 6.

Family of boiling curves for test 3; impingement velocity: 4.8 m/s, water temperature = 40°C. (Online version in color.)

The boiling curve in the parallel flow zone (x = 160 mm) is distinctly different from that in the impingement zone. The transition from air cooling to water cooling occurs at a lower temperature and coincides with the progression of the wetting front. The initial cooling region exhibits a slope much more gradual than in the impingement zone and seems to merge with the transition boiling region at lower temperatures with the absence of a ‘shoulder’. When the temperature drops to 300°C, a broad maximum heat flux region is observed, with values about 1/3rd of those in the impingement zone. The nucleate boiling region below 200°C merges with the boiling curves of the impingement zone.

The boiling curve in the intermediate location appears to show mixed characteristics, with an initial cooling region and shoulder characteristic of the impingement zone, and a subsequent transition boiling regime with maximum showing characteristics of the parallel flow zone, before merging into the nucleate boiling regime. The heat fluxes in this intermediate zone between the two distinct cooling zones show, as a result of the measured temperature fluctuations, substantial fluctuations in the boiling curve.

3.2. Jet Impingement Velocity

Figures 7(a) and 7(b) compare boiling curves in the jet centerline (x = 0 mm) and parallel flow zone (x = 160 mm) for different jet impingement velocities (2.3–4.8 m/s) at a water temperature of 40°C. In the jet centerline (Fig. 7(a)), the heat fluxes in the ‘shoulder’ and the following transition boiling regions increase with increasing jet impingement velocity whereas no influence is observed in the initial cooling and nucleate boiling regions. The maximum heat flux in the boiling curve increases from 14.5 MW/m² to 16.8 MW/m² with an increase in the impingement velocity from 2.3 to 4.8 m/s. For the boiling curves in the parallel flow zone (Fig. 6(b)), heat fluxes in the transition boiling region increase with flow rate, with the maximum heat flux value increasing from ~4 MW/m² to ~6 MW/m² for the investigated range of flow rates. The trend of increasing heat fluxes with increasing impingement velocity is consistent with the observations of Nobari et al.¹¹⁾ for top cooling with a planar nozzle.

Fig. 7.

Comparison of boiling curves for different impingement velocities at (a) x = 0 mm (b) x = 160 mm; (water temperature: 40°C). (Online version in color.)

The maximum heat flux values $(q_{max}^{"})$ for the boiling curves at all eight thermocouple locations are plotted for each impingement velocity in Fig. 8. For all three impingement velocities, the $q_{max}^{"}$ values decrease with increasing distance from the jet centerline. The rate of decrease is moderate in the impingement zone close to the jet centerline (x ≤ 20 mm) before a sharp drop in the $q_{max}^{"}$ values occurs at farther distances (x ≥ 80 mm) before they become nearly independent of distance. This sharp transition of heat flux values delineates the impingement zone from the parallel flow zone.

Fig. 8.

Experimental maximum heat flux with respect to distance from jet centerline for different impingement velocities (water temperature: 40°C). (Online version in color.)

3.3. Water Temperature

Figures 9(a) and 9(b) show the comparison of boiling curves in the jet centerline (x = 0 mm) and parallel flow zone (x = 160 mm) for different water temperatures (10–40°C) at an impingement velocity of 2.3 m/s. Heat fluxes in the jet centerline increase with decreasing water temperatures in the ‘shoulder’ and the following transition boiling regions. The maximum heat flux increases from 14.5 MW/m² to 18.8 MW/m² when the water temperature is decreased from 40 to 10°C. Further, the extent of the ‘shoulder’ in the boiling curves at x = 0 mm becomes wider with increasing water temperature. This trend is in contrast with the observations of Ishigai et al.⁸⁾ for top cooling, where the ‘shoulder’ became narrower at higher water temperatures. Similar to impingement velocity, water temperature does not have a measurable effect on initial cooling and nucleate boiling regions. In the parallel flow zone, the maximum in the boiling curves doubles from ~4 MW/m² to ~8 MW/m² with a decrease in water temperature from 40 to 10°C. The cooling potential of the impinging jet increases with colder water, hence increasing the heat fluxes in a similar fashion as observed by Nobari et al.¹¹⁾ for top cooling. Overall, a more pronounced effect of water temperature is observed in the boiling curves as compared to impingement velocity.

Fig. 9.

Comparison of boiling curves for different water temperatures at (a) x = 0 mm (b) x = 160 mm; (impingement velocity: 2.3 m/s). (Online version in color.)

Figure 10 compares the $q_{max}^{"}$ values at all eight thermocouple locations for different water temperatures. Similar to the results shown in Fig. 8 for different impingement velocities, the $q_{max}^{"}$ values decrease slightly in the impingement zone (x ≤ 20 mm) for all three water temperatures. With increasing distance, the sharp drop in $q_{max}^{"}$ values to the parallel flow zone is less steep with decreasing water temperature, indicating a broader effective impingement zone with colder water. This further confirms the more pronounced effect of water temperature on the $q_{max}^{"}$ values as compared to impingement velocity with increasing distance from the jet centerline.

Fig. 10.

Experimental maximum heat flux with respect to distance from jet centerline for different water temperatures (impingement velocity: 2.3 m/s). (Online version in color.)

3.4. Nozzle Inclination

Tilting the nozzle at an angle causes asymmetry in the flow^13,19) such that two directions, i.e. +x direction (along the inclination) and −x direction (opposite to the inclination), have been defined with respect to the jet centerline, as illustrated in Fig. 3(b). To analyze the influence of nozzle inclination on flow symmetry, cooling curves at equidistant locations were studied for different nozzle orientations (Fig. 11). With no nozzle inclination, the cooling curves are symmetrical on either side of the jet centerline (Fig. 11(a)), with the progression of the wetting front coinciding at equidistant locations (± x). For an inclined nozzle, the cooling curves closer to the jet centerline (±20 mm) in the impingement zone appear to be symmetrical (Figs. 11(b) and 11(c)). At a farther distance (± 60 mm), however, the wetting front progresses at different rates on either side of the jet centerline. The degree of asymmetry increases with nozzle inclination.

Fig. 11.

Cooling curves at different equidistant locations for (a) no nozzle inclination (b) nozzle inclination of 10° (c) nozzle inclination of 20° (impingement velocity: 2.3 m/s, water temperature: 25°C). (Online version in color.)

Figure 12(a) compares the boiling curves at the centerline for different nozzle inclinations showing not any significant influence of inclination angle. The boiling curves in the parallel flow zone (Fig. 12(b)) do not show any effect of nozzle inclination either, and all three boiling curves are essentially the same within the accuracy of the experiments and analysis. The asymmetry of flow with change in nozzle orientation does not seem to effect the boiling curves in the jet centerline and the parallel flow zone. However, for the experiments with an inclined nozzle, the boiling curves at equidistant locations from the jet centerline have differences in shape as well as heat flux values (Figs. 13(a) and 13(b)). The asymmetry in flow is evident when comparing the heat flux values for the +60 mm and −60 mm positions outside the impingement zone. Heat flux values are higher in the +60 mm position, i.e. in the direction of the inclination.

Fig. 12.

Comparison of boiling curves for different nozzle inclinations at (a) x = 0 mm (b) x = 160 mm; (impingement velocity: 2.3 m/s, water temperature: 25°C). (Online version in color.)

Fig. 13.

Comparison of boiling curves at equidistant locations farther away from jet impingement for nozzle inclination of (a) 10° and (b) 20°; (impingement velocity: 2.3 m/s, water temperature: 25°C). (Online version in color.)

To better quantify the effect of nozzle inclination on heat transfer rates, the cumulative heat extracted during cooling was calculated for different nozzle orientations, following the approach adopted by Chester et al.¹³⁾ For the integration of heat fluxes, a spatial area extending from −60 mm to +160 mm with respect to the jet centerline was considered. A linear interpolation technique was used for heat fluxes at grid locations between thermocouples, coupling the change in heat flux values with distance and temporal progression of the wetting front. Heat fluxes were integrated spatially and temporally to obtain the overall heat extracted as a function of time during cooling for each test. Figure 14 shows the evolution of heat extracted during cooling for different nozzle inclinations. In all three tests the plate is cooled from 700 to 100°C such that the total heat extracted over the cooling period is independent of nozzle inclination. But also in the initial cooling stages only marginal variabilities are recorded with no clear trend in terms of inclination. It has to be noted, however, that the role of table rollers on the hydrodynamics of water flow was not included in the current tests.

Fig. 14.

Comparison of total heat extracted for different nozzle inclinations; (impingement velocity: 2.3 m/s, water temperature: 25°C). (Online version in color.)

4. Conclusions

The objective of this study was to establish a database for stationary bottom jet impingement cooling of a steel plate by a planar nozzle. Systematic experiments were conducted under transient conditions to measure spatial sub-surface temperatures, and the effects of impingement velocity, water temperature and nozzle inclination were quantified in terms of surface heat flux. Boiling curves obtained show a strong effect of distance from the water jet on the heat fluxes demarcating clearly impingement and parallel flow zones. Boiling curves in the impingement zone are characterized by the existence of a “shoulder” in the transition boiling regime, whereas the “shoulder” was absent in the boiling curves of the parallel flow zone. A sharp transition in the maximum heat flux values occurs in the intermittent zone between these two distinct zones and boiling curves are of mixed characteristics. Heat flux values increase with increasing impingement velocity, while decreasing water temperatures enhance cooling capacity of the coolant medium thereby increasing heat flux values. A more pronounced effect of water temperature is seen on the heat extraction rates as compared to that of jet impingement velocity, both in the impingement zone and the parallel flow zone. Further, heat extraction rates for the investigated stationary test conditions are not affected by nozzle inclination up to 20°.

The experimental database for stationary bottom cooling tests with a planar nozzle is complementary to the established database for stationary top cooling experiments by Nobari et al.¹¹⁾ Together, they form an upper benchmark limit for run-out table cooling heat transfer models. For run-out table cooling, also the plate speed and the role of table rollers on heat transfer will have to be quantified with dedicated moving plate tests. Selected moving plate test results are already available from previous studies^28,29,30) but a similar experimental data set has yet to be established for bottom jet impingement cooling.

Acknowledgements

The authors acknowledge the financial support of ArcelorMittal Dofasco Inc. and the Natural Science and Engineering Research Council of Canada. Further, they wish to thank Mr. Gary Lockhart, Mr. Ross McLeod, Mr. David Torok and Mr. Carl Ng for their technical contributions and assistance towards the experimental work of this study.

References

• 1)   V.  Schwinn,  J.  Bauer,  P.  Flüss,  H.-J.  Kirsch and  E.  Amoris: Rev. Metall., 108 (2011), 283. https://doi.org/10.1051/metal/2011054
• 2)   C.  Ouchi: ISIJ Int., 41 (2001), 542. https://doi.org/10.2355/isijinternational.41.542
• 3)   A. A.  Gorni and  J. H.  Dolabela da Silveira: J. ASTM Int., 5 (2008), 1. https://doi.org/10.1520/JAI101777
• 4)   K.  Nishioka and  K.  Ichikawa: Sci. Technol. Adv. Mater., 13 (2012), 23001. https://doi.org/10.1088/1468-6996/13/2/023001
• 5)   D. H.  Wolf,  F. P.  Incropera and  R.  Viskanta: Adv. Heat Transf., 23 (1993), 1.
• 6)   S.  Nukiyama: Int. J. Heat Mass Transf., 9 (1966), 1419. https://doi.org/10.1016/0017-9310(66)90138-4
• 7)   G.  Guedia Guemo,  V.  Prodanovic and  M.  Militzer: Steel Res. Int., 90 (2019), 1800361. https://doi.org/10.1002/srin.201800361
• 8)   S.  Ishigai,  S.  Nakanishi and  T.  Ochi: 6th Int. Heat Transfer Conf., Begell House, Danbury, (1978), 445.
• 9)   T.  Ochi,  S.  Nakanishi,  M.  Kaji and  S.  Ishigai: Multi-phase Flow and Heat Transfer III, Part A, Fundamentals, Elsevier, Amsterdam, (1984), 671.
• 10)   A. T.  Hauksson,  D.  Fraser,  V.  Prodanovic and  I.  Samarasekera: Ironmaking Steelmaking, 31 (2004), 51. https://doi.org/10.1179/030192304225011098
• 11)   A. H.  Nobari,  V.  Prodanovic and  M.  Militzer: Int. J. Heat Mass Transf., 101 (2016), 1138. https://doi.org/10.1016/j.ijheatmasstransfer.2016.05.108
• 12)   A. K.  Mozumder,  M.  Monde,  P. L.  Woodfield and  M. A.  Islam: Int. J. Heat Mass Transf., 49 (2006), 2877. https://doi.org/10.1016/j.ijheatmasstransfer.2006.01.048
• 13)   N. L.  Chester,  M. A.  Wells and  V.  Prodanovic: J. Heat Transf., 134 (2010), 122201. https://doi.org/10.1115/1.4007127
• 14)   K.  Morisawa,  J.  Nakahara,  K.  Nagata,  H.  Fujitomo,  T.  Hama and  H.  Takuda: ISIJ Int., 58 (2018), 140. http://dx.doi.org/10.2355/isijinternational.ISIJINT-2017-383
• 15)   Z. D.  Liu,  D.  Fraser and  I. V.  Samarasekera: Can. Metall. Q., 41 (2002), 63. https://doi.org/10.1179/000844302794406306
• 16)   H.  Robidou,  H.  Auracher,  P.  Gardin and  M.  Lebouché: Exp. Therm. Fluid Sci., 26 (2002), 123. https://doi.org/10.1016/S0894-1777(02)00118-8
• 17)   H.  Robidou,  H.  Auracher,  P.  Gardin,  M.  Lebouch and  L.  Bogdanic: Heat Mass Transf., 39 (2003), 861. https://doi.org/10.1007/s00231-002-0335-6
• 18)   D. E.  Hall,  F. P.  Incropera and  R.  Viskanta: J. Heat Transf., 123 (2001), 911. https://doi.org/10.1115/1.1389062
• 19)   B.  Wang,  Z.  Liu,  B.  Zhang,  Z.  Wang and  G.  Wang: ISIJ Int., 59 (2019), 900. https://doi.org/10.2355/isijinternational.ISIJINT-2018-576
• 20)   N.  Seiler-Marie,  J. M.  Seiler and  O.  Simonin: Int. J. Heat Mass Transf., 47 (2004) 5059. https://doi.org/10.1016/j.ijheatmasstransfer.2004.06.009
• 21)   A. B.  Ahmed and  M. S.  Hamed: Int. J. Heat Mass Transf., 91 (2015), 1273. https://doi.org/10.1016/j.ijheatmasstransfer.2015.07.130
• 22)  ed. by M. S. Loveday, M. F. Day and B. F. Dyson,: Measurement of High Temperature Mechanical Properties of Materials (Proc. Symp.), National Physical Laboratory, London, (1982), 58.
• 23)   G.  Franco,  E.  Caron and  M. A.  Wells: Metall. Mater. Trans. B, 38 (2007), 949. https://doi.org/10.1007/s11663-007-9100-z
• 24)   E.  Caron: Ph.D. thesis, The University of British Columbia, (2008). https://doi.org/10.14288/1.0066982
• 25)   P.  Zhang: M.A. Sc. thesis, The University of British Columbia, (2004). https://doi.org/10.14288/1.0078684
• 26)   S.  Vakili and  M. S.  Gadala: Heat Transf. Eng., 34 (2013), 580. https://doi.org/10.1080/01457632.2013.730412
• 27)   D.  Li,  M. A.  Wells,  S. L.  Cockcroft and  E.  Caron: Metall. Mater. Trans. B, 38 (2007), 901. https://doi.org/10.1007/s11663-007-9091-9
• 28)   A. H.  Nobari: Ph.D. thesis, The University of British Columbia, (2014). https://doi.org/10.14288/1.0167525
• 29)   G. A.  Franco,  M. A.  Wells and  M.  Militzer: Can. Metall. Q., 48 (2009), 197. https://doi.org/10.1179/cmq.2009.48.3.197
• 30)   K. V.  Jondhale,  M. A.  Wells,  M.  Militzer and  V.  Prodanovic: Steel Res. Int., 79 (2008), 938. https://doi.org/10.1002/srin.200806224