DEVELOPING ENVIRONMENTALLY FRIENDLY ROLLING LUBRICANTS ANDREAS JOHNSSON, 1 MARIE EKMAN 2ANDERS JANOLS 3 1 Swerea MEFOS, Box 712, 971 25 Luleå - Sweden 2 Sapa Technology 612 81 Finspång - Sweden 3 SSAB Tunnplåt AB 781 84 Borlänge - Sweden Abstract

Swerea MEFOS pilot mill has been used as a rolling lubricant development tool. Oil in water emulsions and an aqueous solution, by the producer termed a conditional emulsion, were tested where the lubricant formulation was altered to optimize the performance for hot rolling of aluminium. Cold rolling tests were also run on steel coils where the presence of rust protective oil was studied as well as the influence of aged emulsion behaviour compared to fresh emulsion. Finally, Sapa developed a vacuum evaporator system for recovering oil from dumped emulsion generated during aluminium hot rolling, followed by separation, centrifugation and flocculation. The cleaned concentrates were reconditioned for re-use. The recycled and recovered emulsion was tested and compared to the original fresh emulsion in Swerea MEFOS pilot mill with results as good as the fresh emulsion. Keywords: hot rolling, cold rolling, emulsion, pilot mill, recycling 1 INTRODUCTION Increases of productivity and improved product quality are constant goals within the rolling industry. The lubrication formulation is constantly overseen as one of the parts of the puzzle to meet the new higher goals. However, it is a costly and not always a problem free process to make industrial on-site tests of lubricants. A more cost efficient way is to test the lubricants on a pilot scale first. The advantage is that there are good probabilities to isolate the lubricant performance and compare the outcome with the performance of other lubricants under controlled conditions. During the pilot rolling mill trials several products were tested. A part of the work was aimed at optimizing the performance of a traditional emulsion and investigate the importance of the various components. Different formulations were tested where additive levels, emulsifier levels, HLB values and viscosity were varied to test the performance of the traditional oil-in-water emulsion. The influence of contamination of the rolling lubricant with rust protective oil to a concentration of up to 10% were tested along with tests comparing an industrially run or aged emulsions performance to a fresh emulsion. Today’s cold-rolling mills meet today’s environmental demands regarding chemical compounds emitted to the environment. It can also be expected that these demands will increase in the future; it is therefore important to find ways to recover and recycle the oil and water phases from traditional oil-in-water emulsions and minimize emissions to the

4th International Conference on Tribology in Manufacturing Processes - ICTMP 2010

environment. Natural resources must also be used more efficiently, so we can retain the production rate using less energy and fewer raw materials. Sapa Heat Transfer have built a vacuum evaporator system designed to recover and clean the oil from dumped emulsions from their industrial plant for hot rolling of aluminium. It was designed to recycle, clean and recover the oil and water for re-use. It consists of a vacuum evaporator, collection tanks and a centrifugal separator. The water, which is a condensate from the evaporation process is clean with a low metal content, goes back to the emulsions system for re-use. The recovered oil-concentrate was cleaned by centrifugation and flocculation, evaluated, stored, re-evaluated and reconditioned for pilot mill rolling. In looking for alternatives to oil-based lubricants, conditional emulsions have a great potential. The conditional emulsion acts as a water solution under normal circumstances but under certain conditions it transfers into behaving as an emulsion. The condition is a certain temperature that is met inside the roll-gap where the pressure and temperature rises to higher levels. After rolling, when the temperarture falls the emulsion transfers back into behaving as a water solution again. It was therefore important to include this new kind of oil-free lubricant in this project. The conditional emulsion, novel to rolling conditions, was optimized in Swerea Mefos pilot mill. The active components and the cloud point were varied and evaluated. The target was set to develop these two alternatives, the recycled and recovered oil-inwater emulsion and the conditional emulsion, to a point where their performance is fully comparable to that of the traditional oil-based lubricants. 2 MATERIAL AND EQUIPMENT 2.1

PILOT ROLLING MILL

The tests were conducted with equipment specially designed for testing and evaluation of hot and cold rolling lubricants within the steel and aluminium industries, see figure 1. The pilot rolling mill specifications can be seen in table 1. The mill was set-up in a 1 stand two-high configuration. The roll mill consisted of a pay-off reel equipped with a brake for back tension control, two conductive rollers- one on each side of the millequipped with pulse transducers for reduction control by flow regulation, while strip tension was measured via an electronic unit, and a delivery reel equipped with an engine for forward tension control.

Using a pilot mill as a rolling lubricant development tool

Reduction control by flow regulation

Cooling

Strip tension Strip speed

Strip tension Strip speed

Lubricant

Delivery reel

Rolling force Rolling torque

Pay-off reel

Figure 1: Roll mill configuration The lubricants were applied using a pneumatic membrane pump connected to flat fan nozzles feeding the rolls or roll gap. For the cold rolling of steel strips the lubrication was applied into the roll gap with a flow of 2 l/min per nozzle/side. The lubricant temperature was 50°C and the water used was for comparative reasons the same as it is during industrial rolling. emulsion content was 2.5%. During the hot rolling of aluminium the lubricant was applied onto the rolls due to the high strip temperature and in line with industrial practice. A lubricant flow of 27 l/min was used, 9 l/min onto the rolls and 27 l/min on the delivery side of the rolls for cooling purposes. The water quality used was the same as during industrial rolling, de-ionized water. Back and forward tension, strip and roll temperature, reduction, strip speed, torque and rolling force were computer logged and the times of the lubricant changes were registered.

Roll mill specifications Roll force

500 kN

Roll torque

2800Nm

Roll speed

0-10 m/s

Rolls 2 high

Ø 160mm, face length 125mm

Payback reel

5kN

Delivery reel

10kN Table 1: Roll mill specifications

4th International Conference on Tribology in Manufacturing Processes - ICTMP 2010

2.2

MATERIAL

The aluminium strips used were delivered by Sapa Heat Transfer. The strip thickness was 3 mm and strip width 90 mm. The tests were performed at a strip temperature of 400°C. A bell type furnace was used for heating up the strip to a homogeneous temperature of 400°C before placing it on the pay-off reel. The steel strips were delivered by SSAB Tunnplåt, Borlänge, with a width of 45mm and a thickness of 2.5mm. The steel was delivered in a hot rolled and pickled condition. Two grades were used, one soft and one hard to roll grade. 3 RESULTS 3.1

CONTAMINATION BY RUST PROTECTIVE OIL AND AGED EMULSION VS FRESH EMULSION. STEEL ROLLING.

Cold roll tests on steel strips were run at Swerea Mefos pilot mill, where SSAB Borlänges previously used emulsion was compared to their currently used rolling emulsion. The emulsions were either contaminated with rust protective oil or aged. The contaminated emulsions were contaminated up to the maximum amount that occurs at their plant, which is 10%. How this contamination influences the rolling process and the rolling emulsion properties was unclear and the focus of the tests. The results show that the rolling force increases most with contamination by rust protective oil for one of the emulsions, the previously used emulsion, but the difference compared to un-contaminated emulsion is not large, see table 2. For the currently used emulsion the contamination did not affect the rolling process to any significant amount. During the ageing tests, only the currently used emulsion was used. The aged emulsion was industrially run before the tests in the pilot mill. Back and forward tensions were fixed during the tests, and roll speed and reductions were varied. The emulsion particle size was measured for a number of tests to comply with industrial conditions. A hard and a soft steel grade was subject to rolling, where focus was on the tougher conditions i.e. the harder to roll grade. The results show that the aged emulsion does not perform as well as the fresh emulsion and the difference increases with reduction and contact pressure, see figure 2.

Increase in rolling force (%)

Reduction (%)

Speed (m/s)

1.8

15

0.1

3.3

30

0.1

0.4

15

2.5

1.2

30

2.5

Table 2: The influence of rust protective oil

Using a pilot mill as a rolling lubricant development tool

Roll force at different settings- new and industrially aged emulsion 450

roll force, kN

400 350 300 250 200 150 speed 0.1m/s red 15%

speed 2.5m/s red 15%

speed 0.1m/s red 30%

speed 2.5 m/s red 30%

speed 2.5 m/s red 30%

setting aged emulsion, hard steel grade fresh emulsion, hard steel grade

aged emulsion, soft steel grade fresh emulsion, soft steel grade

Figure 2: Fresh emulsion versus an industrially aged emulsion tested on a hard to roll material DP800 and a soft material B500 3.2

OPTIMIZATION OF A TRADITIONAL OIL IN WATER EMULSION AND A CONDITIONAL EMULSION. ALUMINIUM ROLLING

Hot rolling was performed on aluminium strips. The total amount of strips rolled was 20. On each strip several lubricants were sequentially applied by switching the lubricant source while keeping rolling conditions constant, such as strip speed, reduction, strip back and forward tensions etc. By using this approach, the output in terms of rolling force and rolling torque can be compared between the different applied lubricants. During the pilot rolling mill trials several lubricants were tested. The tests were aimed at optimizing the performance of a traditional emulsion as well as a new lubricant, in this case a conditional emulsion. Different formulations were tested where additive levels, HLB values and viscosity were varied to test the performance of the traditional lubricant. The droplet size of every lubricant to be used was checked before each strip was rolled, and manual stirring was continued until the correct size was achieved. The active components and the cloud point of the conditional emulsion were also varied and evaluated. De-ionized water was used for the preparation of the emulsions. The tests were performed within or towards the boundary region. This was ensured by running the mill at a constant low strip exit speed of 50 m/min and a high reduction, 65%, which is the same as the industrial case. The roll surfaces were grinded to a surface roughness of 1.1-1.2µm. Back tension during the tests was 3.7 N/mm2 and forward tension 27 N/mm2. A reference test was made, with the same lubricant being applied throughout the test. The reference test had to be performed to provide information about the representative

4th International Conference on Tribology in Manufacturing Processes - ICTMP 2010

part of the strip, or the part where the responding roll force measured will give a fair value. Since the roll force is sensitive to changes in strip temperature, the sensitiveness had to be checked. The strip was heated to a homogenous temperature of 400°C, but the outer rounds of the coiled up strip will be a bit cooler due to test preparation, as will the inner rounds in contact with the pay-off reel. The reference tests show that the part where valid comparisons can be made between different applied lubricants, and where rolling conditions are constant, is between 150 s after the correct test conditions have been reached and 600 s ahead. The results from the pilot mill tests can be seen in figures 3 and 4. Figure 3 show the results from a strip rolled with a traditional emulsion were the formulation was varied with high and low amounts of important components and figure 4 shows a test of the conditional emulsion where the active components and cloud point were varied. The conditional emulsion consists of three major components, which were varied during the tests. The total amount of the components was kept constant, but the relative portion of each component was varied. Step by step the formulations were altered until the best performing formulation was found. The lubricant performance is based on the change in rolling force and torque needed to achieve the same reduction under the same rolling conditions. The traditional oil in water emulsion, which is a commercial lubricant, was optimized and the altered formulation was improved by about 15% compared to the commercial formulation. The conditional emulsion, which was optimized step by step, was improved from low performing with very high rolling forces compared to the traditional emulsion, to performing slightly better then the traditional emulsion. Analysis of surface finish and chemical residues also show good results. Roll separating force and torque vs time 4

250

Roll force, kN

Low Ester

Reference 150

High Viscosity

High Ester

3

2,5 100

Torque, kNm

3,5

200

2 50

0 200

1,5

300

400

500

600

700

800

900

1000

1100

1 1200

time, s Roll force, kN

Torque, kNm

Figure 3: Different formulations of the same emulsion tested sequentially on the same strip. The arrows indicate times of change of lubricant source

Using a pilot mill as a rolling lubricant development tool

Roll separating force and torque vs time 400

3

350

2,5 2

250 1,5 200 1 150

Torque, kNm

Roll force, kN

300

0,5

100

0

50 0 0

200

400

600

800

1000

1200

-0,5 1400

time, s Rolling force, kN

Rolling torque, kNm

Figure 4: Different compositions of the same conditional emulsion tested sequentially on the same strip. The arrows indicate times of change of lubricant source 3.3

RESULTS FROM THE TESTS WITH RECOVERED AND RECONDITIONED ROLLING EMULSIONS. ALUMINIUM ROLLING.

The next step was to investigate the influence of various rolling conditions. The lubricant formulation was kept constant while back and forward tensions were varied to study the neutral plane angle and forward slip. Four different setting were tested 1)high back-high forward tension, 2)low-high, 3) low-low and 4) high-low . A high value was set as 150% the normally used industrial tension, and a low value 50% of the normal tension. Three different reductions were tested 35%, 50% and 65%. The recovered and reconditioned emulsions were tested according to the same schedule as above, with varied rolling conditions. The same emulsion, in the new condition, was thus compared to the recovered and reconditioned emulsion industrially used until it would normally be dumped. The recovered oil concentrate was prepared with two different flocculants, and it was of interest to include both flocculants in the tests. One flocculant was azelaic acid and the other a polymer based flocculant. The test results can be seen in figure 5 and 6.

4th International Conference on Tribology in Manufacturing Processes - ICTMP 2010

Roll force per lubricant and setting 300 65% reduction

Roll force, kN

250

200 50% reduction

150 35% reduction

100

50

0 1

2

3

4

1

2

3

4

1

2

3

4

setting

azelaic acid

polymer

new emulsion

Figure 5: Rolling force at various reductions and tensions for the fresh emulsion and the recovered emulsion flocculated with each of the two flocculants.

Rolling torque per lubricant and setting 2,5 65% reduction

Roll torque, kNm

2

1,5

50% reduction 35% reduction

1

0,5

0 1

2

3

4

1

2

3

4

1

2

3

4

setting azelaic acid

polymer

new emulsion

Figure 6: Rolling torque at various reductions and tensions for the fresh emulsion and the recovered emulsion flocculated with each of the two flocculants

Using a pilot mill as a rolling lubricant development tool

As can be seen, the recovered emulsion treated with azelaic acid generates the same rolling force as the fresh emulsion, and lower torque at 65% reduction. The recovered emulsion treated with the polymer based flocculant generates lower rolling force and torque. A likely reason is that the polymer based flocculant generated a somewhat larger droplet size causing better plate-out and thicker oil film. 4 CONCLUSIONS The pilot mill serves very well as a lubricant optimization tool. The traditional oil in water emulsion was optimized in performance by about 15%. The conditional emulsion, which had never been optimized was improved from poor performing to performing even better then the traditioanal emulsion. SAPA is at the moment using the conditional emulsion during regular rolling in the tandem mill. The tests of the recovered and reconditioned emulsion had very good results, and performed as well as a fresh emulsion. All results were backed up by surface analysis and analysis of residuals on the strip surfaces indicating very good results. 5 ACKNOWLEDGEMENTS The work was performed in a project 'Optimization of rolling lubricants for improved operation of cold rolling mills', RFC-CR-04019, that was partly funded by the RFCS. 6 REFERENCE [1]

RFCS Project RFC-CR-04019 ''Optimization of rolling lubricants for improved operation of cold rolling mills".