Modified Jominy Test for Determining the Critical Cooling Rate for Intercritically Annealed Dual Phase Steels [2]

1 of 16

Modified Jominy Test for Determining the Critical Cooling Rate for Intercritically Annealed Dual-Phase Steels

Brendan Connolly

Advisors: A.J. DeArdo C.I. Garcia

2 of 16

ABSTRACTThe Jominy test is a common method of examining the hardenability of steel grades. This test varies the cooling rate along the length of a Jominy bar in order to examine the ability of the steel to form martensite. In this work, the Jominy test has been modified to examine dual phase sheet steels. The reasons for examining the dual phase steels are to measure the critical cooling rates and then empirically determine equations that describe the critical cooling rates accurately. The modified Jominy set-up shows useful results for examining the transformation behavior of austenite in dual phase steels. Alloying elements are the only parameters accounted for in current critical cooling rate equations. The tests showed that the critical cooling rate for the dual phase steels examined will be dependent on more than just composition.

INTRODUCTIONStandard Jominy Test The Jominy test is a standard test used for measuring the hardenability of steels. Figure 1 shows the standard Jominy test sample. The sample is 1 diameter and 4 long. To perform the standard Jominy endquench test, the sample is heated to above the austenitizing temperature and held for one hour (temperatures for common steels are specified by ASTM A-255). The sample is then removed from the furnace and quenched at one end with flowing Figure 1: Standard jominy sample water. In a standard Jominy test, the cooling rate decreases as you move away from the quenched end of the sample. Depending on the composition of the alloy, this decreasing cooling rate may not be fast enough to transform all of the austenite into martensite, and the sample will decrease in hardness along the length of the bar. Once the bar has been quenched, opposite sides are ground or machined to be planar (about in depth), and hardness tests are taken every 1/16 from the quenched end. This gives what is called a hardness curve or H-band. Some typical hardness curves for different steel compositions are shown below in Figure 2.

3 of 16

Figure 2: Typical hardness curves

Dual-Phase Steels Dual-phase steels have generated interest in the automotive industry because of their high strength, high ductility and high work hardening rates1,2,3. Common high strength steels that are used today for automotive application experience poor formability, an area where dual phase steels excel. The dual phase steels have a potential to provide strong but light weight parts for use in transportation and other industries. The weight reduction translates into better fuel economy and therefore more environmentally sound transportation. The increase in strength also allows for excellent dent and crash resistance1. The dual-phase steel considered in this project consists of a ferrite-martensite uniformly mixed microstructure. According to Furukawa2, the martensite is usually preferred to be of the low-carbon type which requires the bulk carbon content to be low (less than 0.10 wt%). In order to achieve the ferrite-martensite microstructure, the material is intercritically annealed at temperatures in the + region. The intercritical anneal temperature determines the volume fraction of austenite that will be formed during annealing. The austenite formed during the intercritical anneal will undergo one or more phase changes upon cooling. The critical cooling rate is the slowest possible cooling rate at which all of the austenite will transform into martensite, and no other phase change will occur.

4 of 16 Determining the critical cooling rate allows for the slowest possible cooling rate that will still achieve full austenite to martensite transformation. A fast cooling rate in industry would require slower line speed and therefore slower production. Figure 3 shows three cooling rates, the intermediate being the critical cooling rate. The compositions of the dualphase steels to be examined in this research are given in Table 1 below.Figure 3: Different cooling rates after IC anneal.

Steel BDP 590 BDP 780

%C 0.06 0.06

%Mn 1.5 1.48

%Si 0.4 0.4

%P 0.01 0.01

%Al 0.06 0.05

%Cr+Mo %Nb 0.48 0.02 0.48 0.04

%N 0.006 0.006

Table 1: Steel compositions to be examined

Critical Cooling Rate Several equations have been developed that allow for calculation of the critical cooling rate based on steel composition. The equations normally use the composition of the steel to determine the shift of the CCT curves. Furthermore, no equations could be found that include the niobium content of the steel, which will affect the hardenability and therefore the critical cooling rate. Many of the equations are based on the manganese equivalent such as: Log (CR) = 3.95 1.73[Mneq] Where Mneq is given by: Mneq = A(%Mn) + B(%Si) + C(%Mo) + (Eqn 2)5 (Eqn 1)5

such that A, B, and C are constants. These equations include different elements that were often the main alloying elements used for the steel and the carbon content is normally constant, with the AC3 carbon content neglected. There are also several equations that give a critical cooling rate for formation of the desired dual-phase structure at a given temperature and volume fraction of austenite such as:

5 of 16 Log (CR) = 4.93 1.70[Mn] -1.34[Si] 5.68[C] (Eqn 3)6

In all cases the coefficients to the alloy content are determined by their effect on the hardenability of the steel. In equation 1 the carbon constant is assumed to be constant (0.05%) for the steels used, but in equation three it can be seen that the carbon content has the most effect on the hardenability of the steel. For this reason the carbon content is calculated for the experiments to be performed. There are also equations that relate the cooling rate to the ultimate tensile strength and composition 4, but the ultimate tensile strength can be affected by so many other parameters that these equations will not be used for this research. After the intercritical anneal the microstructure consists of ferrite (both recrystallized and non-recrystallized) and austenite. The austenite, upon cooling at the critical cooling rate will avoid transformation into other phases and will transform into martensite. The volume fraction of austenite that is formed will be the volume fraction of martensite after cooling. This means that the intercritical annealing temperature is the parameter that will control both the final volume fraction and composition of the austenite prior to transformation. A higher temperature will form more austenite with lower carbon content, while low intercritical temperatures will form less austenite with higher carbon content. Therefore, the intercritical anneal temperature will also change the required critical cooling rate to form martensite by varying the carbon present in the austenite. From this it can be seen that the critical cooling rate will be dependent on composition as well as intercritical annealing temperature.

EXPERIMENTModified Jominy Cold-rolled sheet steels (~2mm in thickness) can not be tested using the standard Jominy test. The introduction of cold work gives a higher potential for ferrite recrystallization. ASTM A-255 specifies that Jominy samples that have undergone cold work prior to machining are to be normalized before the Jominy test is performed. In order to examine the effect of different cooling rates (and thereby determine the critical cooling rate) on dual-phase steel sheets, the sheet would ideally be in the cold-rolled condition during testing. Modifications to the standard Jominy test were done to allow the examination of dual-phase sheet steels under many different cooling rates all within one sample. One type 420 martensitic stainless steel Jominy specimen was cut in half along its length. Approximately 2 mm of material was removed during the cutting process. Six recessed holes were drilled into one half of the Jominy sample, and six smaller holes were drilled and tapped to accept type 632 screws on the other half. The schematic used by the machine shop is shown in Figure 4.

6 of 16

Figure 4: Schematics used for machining

This modified Jominy tests allows for examining many different cooling rates after an intercritical anneal using only one sample. The modified Jominy test was first performed on a piece of 4140 steel. This was used to compare literature hardness values to our experimental hardness values. The 4140 material was machined to a planar 1x4x2mm piece. Six holes were drilled in the sample to match those of the connecting screw holes. The sample was heated to 1550F (as specified by ASTM) and held for 1 hour. The sample was then quenched and hardness values were taken along the length from the quenched end. A second experiment was performed to determine the cooling rates at different positions within the sample. Three k-type thermocouples were connected to a Lab View data acquisition program to monitor temperatures within the sample during quenching. The thermocouples were 1.5 mm in diameter, inserted into holes drilled in the sample at 5/16, 1 , and 2 5/8 from the quenched end. The holes were drilled as shown in Figure 5.

7 of 16

Figure 5: Thermocouple positions in sample

Temperatures were taken every second from each thermocouple. Dual-Phase Experiments The tests were performed using three different heat-treatment temperatures for each of the dual phase compositions. Two of the temperatures fall in the intercritical region (745C and 785C) and one temperature that should give 100% austenite (860C). These temperatures were calculated with the aid of J-Mat Pro. The 745C intercritical anneal was predicted to give ~20% austenite with a carbon content of 0.27% carbon. The 785C intercritical anneal was predicted to give approximately 40% austenite at 0.14% carbon. The supplied dual phase steel sheets each had several 1x 4 sections cut from them. Each section had holes drilled to match the arrangement of holes in the modified Jominy. A high temperature lubricant was used on the connecting screws to ensure that they could be easily removed at the end of each experiment. Hardness tests were performed on each sample after quenching. Tests were performed using a standard Rockwell B scale. Dual-phase samples were cut to examine microstructures at cooling rates of 30C/s, 10C/s, and 5C/s (cooling rates at 700C). The quenched end microstructures were also examined (cooling rate of greater than 200C/s). The locations of these cooling rates were determined using Figure 8. All samples were mounted, polished, and etched with 2% nital. The samples were examined on both a PSEM (Personal SEM) and a Philips XL-30 field emission scanning electron microscope. Volume fraction measurements were done using a 120 point grid.

RESULTS & DISCUSSIONThe hardness values obtained from the 4140 modified Jominy experiment were compared to ASTM upper and lower hardness limits for 4140 steel in a standard Jominy test. The values of the experimental data, and upper and lower ASTM limits are given below in Figure 6.

8 of 16

Hardness curves - 414080.0 60.0 HRC 40.0 20.0 0.0 0 5 10 15 20 25 30 35 40 45 distance from quenched end (1/16")Figure 6: Hardness curve for experimental 4140

avg high lim low lim Poly. (avg)

Figure 6 shows that only one point is outside of the ASTM specification. This point is at a distance where holes were drilled in the sample to accommodate the connecting screws. The data obtained from the sample with embedded thermocouples is given in Figures 7 and 8. When compared to work done by Wongpiromsarn7 on standard Jominy cooling rates, the cooling rates near the quenched end in the modified Jominy were much higher. This was most likely caused by thermal contraction and expansion of the pieces of the modified Jominy. When the modified Jominy is heated, the pieces expand, once they begin to cool during quenching, the pieces contract at the quenched end and leave a small gap between the sample and the modified Jominy. However, this should not affect the validity of the data as long as the cooling rates are known and consistent. The type of curve seen for the cooling rates vs. distance from quenched end (Figure 8) is the same for this experimental work and the work done by Wongpiromsarn7. This further validates that the data obtained from the thermocouples is accurate.

9 of 16

Time vs. Temp1000 Temp (C) 800 600 400 200 0 0 200 400 600 Time (s)Figure 7: Temperature-Time curves for thermocouple in sampleJ5 Thermocouple

J24 Thermocouple

J42 Thermocouple

Cooling rate vs distance from quenched end80.0 70.0 Cooling rate (C/s) 60.0 50.0 40.0 30.0 20.0 10.0 0.0 0 10 20 30 40 50 Distance from quenched end (1/16")Figure 8: Cooling rates (at 700C) for sample calculated using thermocouple data.

Literature data Experimental data Power (Literature data) Power (Experimental data)

10 of 16 The hardness tests did not show large changes in hardness over the length of the specimen. Only the samples heat treated at 860C showed a significant decrease in hardness from the quenched end to the slow-cooled end. The samples containing 0.04% Nb (BDP 780) did show higher hardness values than the samples containing 0.02% Nb (BDP-590). A table of hardness values is given in Table 2. The J-position values denote the distance from the quenched end in 1/16.HRB JPOS 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 18 20 22 24 26 28 30 32 34 36 38 40 860 104 97 96 92 91 90 89 89 89 89 88 88 88 88 88 88 87 88 88 88 87 87 85 85 85 85 NA NA BDP 780 - 0.04 Nb 785 93 93 94 93 93 93 93 93 93 93 93 93 93 93 93 93 92 92 92 92 91 92 86 86 90 91 92 91 745 90 90 90 89 89 89 89 90 89 89 89 88 88 90 89 89 89 88 88 89 89 89 86 86 89 90 90 90 860 101 95 94 93 92 92 91 90 89 88 88 88 88 88 88 88 88 86 86 86 85 85 78 78 83 85 86 85 BDP 590 - 0.02 Nb 785 87 87 87 88 90 90 91 91 90 90 89 89 90 90 89 89 88 88 89 89 88 88 83 83 88 87 88 88 745 84 83 84 84 84 84 84 84 83 84 83 84 84 84 85 84 84 85 84 84 84 83 76 80 83 83 78 79

Table 2: Hardness data from dual-phase modified jominy test

The micrographs showed the general trends that were expected: Decreasing amount of martensite further from quenched end More martensite present in samples with higher heat-treatment temperatures More uniform martensite island size in samples with higher niobium content

There was significantly less martensite formed in the experiments than predicted by JMat Pro. The reasons for this are not clear, but J-Mat Pro is equilibrium thermodynamic modeling software, and there may not have been enough time in the experiments to reach the predicted equilibrium states. Figures 9 32 show the micrographs taken from the SEM. Figures 23 and 24 were taken on the PSEM, an all others were taken using the Philips XL-30 field emission scanning electron microscope.

11 of 16

BDP 60 - 745C intercritical anneal Predicted austenite formation = 20% containing 0.26% carbon

Figure 9: BDP 790, J1 postion (>200C/s), 10%

Figure 10: BDP 790, J8 postion (30C/s), 11%

Figure 11: BDP 790, J14 postion (10C/s), 2%

Figure 12: BDP 790, J22 postion (5C/s), 1%

12 of 16

BDP 60 - 785C intercritical anneal Predicted austenite formation = 40% containing 0.14% carbon

Figure 13: BDP 790, J1 postion (>200C/s), 28%

Figure 14: BDP 790, J8 postion (30C/s), 13%

Figure 11: BDP 790, J1 positi Figure 11: BDP 790, J1 positi

Figure 15: BDP 790, J14 postion (10C/s), 13%

Figure 16: BDP 790, J22 postion (5C/s), 1%

13 of 16

BDP 60 - 860C intercritical anneal Predicted austenite formation = 100% containing 0.06% carbon

Figure 17: BDP 790, J1 postion (>200C/s), 37%

Figure 18: BDP 790, J8 postion (30C/s), 16%

Figure 19: BDP 790, J14 postion (10C/s), 15%

Figure 20: BDP 790, J22 postion (5C/s), 12%

14 of 16

BDP 61 - 745C intercritical anneal Predicted austenite formation = 20% containing 0.26% carbon

Figure 21: BDP 580, J1 postion (>200C/s)

Figure 22: BDP 580, J8 postion (30C/s)

Figure 23: BDP 580, J14 postion (10C/s), 2000x

Figure 24: BDP 580, J22 postion (5C/s), 2000x

15 of 16

BDP 61 - 785C intercritical anneal Predicted austenite formation = 40% containing 0.14% carbon

Figure 25: BDP 580, J1 postion (>200C/s), 23%

Figure 26: BDP 580, J8 postion (30C/s), 16%

Figure 27: BDP 580, J14 postion (10C/s), 17%

Figure 28: BDP 580, J22 postion (5C/s), 13%

16 of 16

BDP 61 - 860C intercritical anneal Predicted austenite formation = 100% containing 0.06% carbon

Figure 29: BDP 580, J1 postion (>200C/s), 39%

Figure 30: BDP580, J8 postion (30C/s), 11%

Figure 31: BDP 580, J14 postion (10C/s), 4%

Figure 32: BDP 580, J22 postion (5C/s),