Ductile Iron Society

RESEARCH PROJECT No. 43

DIMENSIONAL GROWTH OF DUCTILE IRON CASTINGS

DURING HEAT TREATMENT

BY

RICHARD B. GUNDLACH Stork Climax Research Services Wixom, MI

DUCTILE IRON SOCIETY

Issued by the Ductile Iron Society for the use of its Member Companies – Not for General Distribution

DUCTILE IRON SOCIETY 15400 Pearl Road, Suite 234 Strongsville, Ohio 44136 (440) 665-3686

MARCH 2009

DIS Research Project No. 43

DIS RESEARCH PROJECT NO. 43

DIMENSIONAL GROWTH OF DUCTILE IRON CASTINGS DURING HEAT TREATMENT

ABSTRACT

The dimensional changes upon heat treatment were studied in three grades of ductile iron – ferritic, ferritic-pearlitic and pearlitic grades. The amount of growth was measured in five different heat treatment cycles, including subcritical annealing, full annealing, normalizing and quench and tempering. All materials increased in length upon heat treatment. The dimensional change ranged from less than 0.001 inch per inch to greater than 0.004 inch per inch.

The maximum growth occurred in annealing, and the amount of growth in annealing was proportional to the amount of pearlite in the starting microstructure. Significantly less growth occurred in hardening operations, that is, in normalizing or quench and temper operations. The results indicate that the starting microstructure influences the amount of growth that occurs in quenching and tempering. In addition, more growth occurred at the higher tempering temperature. The increased growth at the higher tempering temperature is attributed to graphitization during tempering.

The amount of growth that occurs is complex due to several reactions that can occur during the heat treatment cycle. Initially, partial decomposition of pearlite and graphitization on heating to the austenitizing temperature can occur. Additional expansion may occur due to re-carburization of the matrix during austenitizing.

Page 2 of 20 DIS Research Project No. 43

Research Project No. 43

DIMENSIONAL GROWTH OF DUCTILE IRON CASTINGS DURING HEAT TREATMENT

INTRODUCTION When ductile iron castings are heat treated, there is a dimensional change that occurs. The dimensional change is primarily caused by (1) a change in the density of the metallic matrix and (2) a change in the volume occupied by the graphite nodules. Most often the castings grow during heat treatment. There are reports that the change in dimensions of the casting during heat treatment can cause the casting to be out of spec, such that there is insufficient machining stock to meet the finish part dimensions.

The extent of growth is directly related to the starting as-cast microstructure. For example, when ferritizing a ferritic-pearlitic casting, the graphitization of pearlite produces growth. The amount of growth that occurs in the heat treatment of Q&T ductile iron is influenced by the austenitizing temperature and the tempering temperature. The austenitizing temperature influences the amount of carbon that dissolves in the austenite prior to the quench, and thus influences the density of the austenite. Upon quenching, the dissolved carbon content influences the growth and density (lattice parameter) of the martensite that forms. Upon tempering, the tempering temperature influences the amount of contraction that occurs as the martensite lattice gives up carbon and carbides precipitate within the matrix.

The need to know the direction and magnitude of growth that occurs during the heat treatment of ductile iron castings is very important to the foundry and its customers. Both the producer and buyer of ductile iron castings need to have this knowledge when deciding if and what heat treatments are acceptable. This is particularly true when heat treatment is offered as a remedy to salvage castings.

While it is readily acknowledged that ductile iron castings grow during heat treatment, no systematic study has been conducted to determine the magnitude of growth in such a broad range of materials and heat treatments.

OBJECTIVE The objective of this program was to determine the dimensional changes, or growth, that occurs in ductile iron castings during various heat treatments -- both subcritical and super critical heat treatments. Some common heat treatments applied to ductile iron will be investigated including ferritize annealing, normalizing, and quench & tempering. The dimensional changes will be determined for ferritic, ferritic-pearlitic and pearlitic ductile iron materials.

Page 3 of 20 DIS Research Project No. 43 PROCEDURES AND RESULTS The materials used for this study included ferritic, ferritic-pearlitic and pearlitic grades of ductile iron. Cast test bars of these materials were provided by Neenah Foundry and the compositions and mechanical properties are shown in Table 1. A total of 5 heat treatment cycles were employed in this study as presented in Table 2 below.

Table 1 List of cast materials provided by Neenah Foundry for this study

SPECIMEN D65M D80 D100 C 3.68 3.57 3.67 S 0.004 0.004 0.003 P 0.024 0.029 0.022 Si 2.44 2.47 2.41 Mn 0.25 0.28 0.26 Cu 0.09 0.46 0.98 Sn 0.004 0.004 0.005 Cr 0.047 0.043 0.039 Mo 0.005 0.007 0.004 Ni 0.03 0.02 0.02 V 0.005 0.006 0.006 Al 0.018 0.016 0.018 Ti 0.02 0.022 0.019 Mg 0.04 0.035 0.039 Ce 0.004 0.004 0.005 Pb 0.0005 0.0005 0.0002 Sb 0.0004 0.0005 0.0005

Tensile 65.5 91.8 120.9 Strength, ksi Yield 44.1 56.1 72.7 Strength, ksi Elongation, % 18.9 11.5 5.6 Hardness, HBW3000 163 217 269

Nodularity, % 92.6 94.2 94.3 (by count) Nodularity, % 91.2 93.1 94 (by area) Nodule Count 276 300 211 Pearlite, % 6.7 50 84.4 Ferrite,% 78.7 38.9 6.1

Page 4 of 20 DIS Research Project No. 43 Table 2 Materials and Test Matrix

Heat Treatment to be Initial Condition of Casting Evaluated Ferritic F + P Pearlitic Ferritize Ann. Supercritical X X Ferritize Ann. Subcritical X X Normalize (pearlitic) X X Q&T 300 HB X X X Q&T 450 HB X X X

Growth Measurement The experimental procedure consists of collecting ductile iron materials for this study, followed by machining test specimens for the heat treatment and the measurement of dimensional change. For the annealing cycles, a dilatometer utilizing 25 mm long specimens was employed. The dilatometer allowed the measurement of linear expansion and contraction throughout the heating and cooling cycle.

For those thermal cycles requiring rapid cooling, longer (50 mm) specimens were machined and the length of each specimen was determined before and after heat treatment using precision measurement instrumentation.

As mentioned above, the austenitizing temperature, cooling rate, and tempering temperature have significant influences on the total growth of ductile iron. Care was taken to employ temperatures and times typically used in the heat treatment of ductile iron. Therefore, the heat treatment parameters used in this study are shown in Table 3. The results of the measurements of change in length for the growth specimens are given in Table 4. Dilatometer records for the annealing cycles of the D5506 and D7003 materials are shown in Figures 1 to 4

Table 3 Heat Treatment Parameters used in this study

Heating Cycle Soak Cooling Heat Treatment Time Rate Temperature Time Full Anneal 872oC 1.5 h 1.5 h 3oC/min Subcritical Anneal 718oC 1.5 h 1.5 h 3oC/min Normalizing 900oC 1.5 h 1.5 h Air Quench & Temper 872oC 1.5 h 1.5 h Oil (450 HB) Quench & Temper 872oC 1.5 h 1.5 h Oil (300 HB)

Page 5 of 20 DIS Research Project No. 43 Table 4 Results of Linear Growth Measurements in Three Ductile Iron Alloys

Sample Temperature Length Change in Change in Condition ID oF mm length, % length, in/in

100A Quench & Temper 1600/950 50 0.122 0.0012 in 80A Quench & Temper 1600/950 50 0.207 0.0021 in 65A Quench & Temper 1600/950 50 0.080 0.0008 in

100B Quench & Temper 1600/1100 50 0.142 0.0014 in 80B Quench & Temper 1600/1100 50 0.178 0.0018 in 65B Quench & Temper 1600/1100 50 0.134 0.0013 in

80C Normalized 1650 50 0.177 0.0018 in 65C Normalized 1650 50 0.068 0.0007 in

100A Subcritical Ann 1325/4h 25 0.430 0.0043 in 80A Subcritical Ann 1325/4h 25 0.288 0.0029 in

100B Full Anneal 1600/1.5h 25 0.323 0.0032 in 80B Full Anneal 1600/1.5h 25 0.333 0.0033 in

Metallography Samples were removed from the as-received test bars and mounted for metallographic examination. Following the growth study, samples were also removed from each of the growth specimens. All samples were compression-mounted in Bakelite, then ground and polished on silicon carbide and diamond abrasives. The polished specimens were etched in 4% picral or 2% nital to reveal the matrix microstructure. Representative photomicrographs of the three as-received materials are shown in Figures 5 to 7. The microstructures of the heat-treated growth specimens are shown in Figures 8 to 19.

DISCUSSION The dimensional changes upon heat treatment were studied in three grades of ductile iron – ferritic, ferritic-pearlitic and pearlitic grades. All materials increased in length upon heat treatment. The dimensional change ranged from less than 0.001 inch per inch to greater than 0.004 inch per inch.

Annealing The dimensional growth occurring during annealing was as anticipated for ductile iron materials containing pearlite. Large amounts of growth occurred during subcritical annealing due to the decomposition of pearlite and the formation of additional graphite. Graphite is less dense than the iron carbide phase and, thus, the specimens expanded. The curves in Figures 1 and 2 show the large expansions that occurred on heating at 718C, which is below the lower critical temperature. The amount of growth was

Page 6 of 20 DIS Research Project No. 43 proportional to the amount of pearlite, with the grade D7003 exhibiting more growth than the grade D5506 material. The results of growth during annealing are summarized in Table 5.

In full (super critical) annealing, even more growth occurred in the grade D5506 material. The increased growth over subcritical annealing is attributed to two possible stages of growth. During heating to the lower critical temperature, partial decomposition of the pearlitic constituent results in significant graphitization and growth. This growth is visible in the dilatometer record shown in Figure 20. Upon heating above the critical temperature, the matrix becomes recarburized by the graphite as the solubility of carbon in the austenite phase is quite high. Upon slow cooling through the critical temperature range and below, fresh graphite is formed and additional growth occurs. For the grade D5506 material, the full anneal cycle produced growth of 0.33% whereas subcritical annealing produced only 0.29%.

In full annealing of the grade D7003, there was less growth than in the subcritical annealing cycle. This is attributed to the fact that the full-annealed specimen was not completely converted to 100% ferrite – it still contained substantial amounts of pearlite. Like the D5506, the D7003 material displayed partial decomposition on heating to the critical temperature, as shown in the dilatometer record in Figure 21. However, upon cooling from the austenitizing temperature some pearlite formed and less graphitization (expansion) occurred, as shown in the micrograph of the supercritical annealed sample in Figure 9. As a result, less growth occurred in the full anneal cycle than that in the subcritical anneal cycle. It is surmised that, had a slower cooling rate been employed, more ferrite (and graphite) would have developed and greater growth would have occurred.

Normalizing Upon normalizing the D4512 and D5506 materials, both exhibited dimensional growth. The grade D5506 material displayed greater growth than the grade D4512 material. No dilatometer records were available for this cycle. It is surmised that the increased growth of the D5506 is, once again, due to partial decomposition of the pearlite during heating to the lower critical temperature. Once reaching the upper critical, both alloys would experience recarburization and only modest growth would occur on air cooling, when transformation to pearlite occurred. The pearlitic transformation involves little graphitization, and the amount of expansion from the austenite-to-pearlite reaction produces much less dimensional change than occurs on annealing, when austenite decomposes to ferrite and graphite.

Quenching and Tempering Upon quenching and tempering, the amount of dimensional change in both grades D5506 and D7003 are similar to that in normalizing grade D5506. The similarity in growth is attributed to the largely pearlitic microstructure in both materials, as can be seen in the as-received microstructures of these materials in Figures 6 and 7.

As in normalizing, partial decomposition of the pearlite constituent occurs during heating to the lower critical temperature. Decomposition of the pearlite to ferrite and graphite produces some expansion. Upon further heating above the critical temperature, the alloys become recarburized. Upon quenching from the austenitizing temperature, Page 7 of 20 DIS Research Project No. 43 martensite forms and a large amount of expansion is anticipated. Indeed, the martensitic reaction in high-carbon austenite is quite large and can lead to quench cracking. However, on tempering the carbon leaves the lattice, the martensite lattice contracts, and the resulting microstructure consists of ferrite and carbide. The end result is that the constituents (and density) of tempered martensite approach those of pearlite. The extent to which the density of tempered martensite approaches that of pearlite depends on the time and temperature of the tempering cycle.

It is noteworthy that the dimensional change of the ferritic D4512 material is less than that of the pearlitic grades, particularly when tempered at 950F. This is attributed to the fact that the ferritic grade lacked significant amounts of pearlite and, therefore, it did not experience graphitization and growth during heating to the lower critical temperature.

On tempering at 1100F (to obtain lower hardness and increased ductility and toughness), both grades D4512 and D7003 experienced greater dimensional change than when tempered at 950F. The increased dimensional growth that occurred on tempering at 1100F is attributed to further graphitization of the martensite. It is well known that secondary graphitization occurs in quench and tempered ductile iron at temperatures of 950F and above. The added graphitization produces more growth. Secondary graphitization was particularly notable in the microstructure of the ferritic grade, see Figure 17.

Table 5 Summary of growth and the amount of pearlite

% Linear Grade % Pearlite Growth In Subcritical Annealing: D7703 0.43% 84.4 D5506 0.29% 50 In Supercritical Annealing: D7703 0.32% 84.4 D5506 0.33% 50 In Normalizing: D4512 0.068% 6.7 D5506 0.177% 50

Growth and Decomposition of Pearlite The graphitization associated with pearlite decomposition contributed significantly to the growth of the alloys containing pearlite. Table 5 lists the growth observed in subcritical annealing, in supercritical annealing, and in the normalizing heat treatment cycles. The amount of growth associated with the decomposition of pearlite on heating to the lower critical temperature was substantial, as shown in Figures 20 and 21.

Page 8 of 20 DIS Research Project No. 43 In this study, there was an effort to maintain a relatively constant rate (10 deg. C/min.) of heating to the lower critical temperature in order to obtain consistent growth values. If slower heating rates had been applied, it is anticipated that the amount of growth occurring in Grades D5506 and D7003 would have been greater. Likewise, had the heating rates been faster, the amount of growth in these grades would be expected to be lower.

SUMMARY The results of this study indicate that all five heat treatments resulted in dimensional growth in all three grades of ductile iron. The results also show that the starting microstructure influences the amount of growth that occurs, whether annealing, normalizing or quenching and tempering.

The primary cause for growth is attributed to graphitization. Graphitization causes growth because graphite and ferrite are less dense than the pearlite from which they form. One source of growth occurred when pearlite decomposed on heating to the lower critical temperature. Further graphitization occurred when martensite decomposed with more growth occurring at the higher tempering temperature.

There appear to be six reactions that contribute to the growth of ductile iron parts during heat treatment. They are as follows.

• Graphitization of pearlite • Recarburization of austenite • Secondary graphitization of martensite • Irreversible growth due to graphitization • Martensite formation • Tempering of martensite

RECOMMENDATIONS The results of this project indicate that a significant amount of growth occurs on heating to the austenitizing temperature due to partial decomposition of pearlite. The amount of growth will be dependent on the amount of pearlite present, and on the rate of heating. It is recommended that further study be conducted to better establish the relationships between pearlite content, heating rate and growth in ductile iron castings.

ACKNOWLEDGEMENTS The test bars received from Neenah Foundry for this study are gratefully acknowledged.

STORK CLIMAX RESEARCH SERVICES Richard B. Gundlach Senior Metallurgical Engineer

Page 9 of 20 DIS Research Project No. 43

Figure 1 Dilatometer record for grade D5506 in the subcritical anneal cycle.

Figure 2 Dilatometer record for grade D7003 in the subcritical anneal cycle.

Page 10 of 20 DIS Research Project No. 43

Figure 3 Dilatometer record for grade D5506 in the full anneal cycle.

Figure 4 Dilatometer record for grade D7003 in the full anneal cycle.

Page 11 of 20 DIS Research Project No. 43

Figure 5. Ferritic D4512 ductile iron as-received. Mag. = 100X

Figure 6. Ferritic-pearlitic D5506 ductile iron as-received. Mag. = 100X

Page 12 of 20 DIS Research Project No. 43

Figure 7. Pearlitic D7003 ductile iron as-received. Mag. = 100X

Page 13 of 20 DIS Research Project No. 43

Figure 8. Ferritic-pearlitic Grade D5506 following full anneal. Mag. = 100X

Figure 9. Pearlitic Grade D7003 following full anneal. Mag. = 100X

Page 14 of 20 DIS Research Project No. 43

Figure 10. Ferritic-pearlitic Grade D5506 following subcritical anneal. Mag.= 200X

Figure 11. Pearlitic Grade D7003 following subcritical anneal. Mag. = 200X

Page 15 of 20 DIS Research Project No. 43

Figure 12. Ferritic Grade D4512 following normalizing. Mag. = 200X

Figure 13. Ferritic-pearlitic Grade D5506 following normalizing. Mag. = 200X

Page 16 of 20 DIS Research Project No. 43

Figure 14. Ferritic Grade D4512 following Quenching & Tempering at 950F. Mag.= 500X

Figure 15. Ferritic-pearlitic Grade D5506 following Quenching & Tempering at 950F. Mag.= 500X

Page 17 of 20 DIS Research Project No. 43

Figure 16. Pearlitic Grade D7003 following Quenching & Tempering at 950F. Mag.= 500X

Figure 17. Ferritic Grade D4512 following Quenching & Tempering at 1100F. Mag.= 500X

Page 18 of 20 DIS Research Project No. 43

Figure 18. Ferritic-pearlitic Grade D5506 following Quenching & Tempering at 1100F. Mag.= 500X

Figure 19. Pearlitic Grade D7003 following Quenching & Tempering at 1100F. Mag.= 500X

Page 19 of 20 DIS Research Project No. 43

Graphitization of Pearlite

Figure 20. Growth during supercritical annealing of the ferritic-pearlitic grade. Note the growth associated with pearlite decomposition on heating to the lower critical temperature.

Figure 21. Growth during supercritical annealing of the pearlitic grade. Note the growth associated with pearlite decomposition on heating to the lower critical temperature. Page 20 of 20

DUCTILE IRON SOCIETY