VACUUM FURNACE HEAT TREATING AEROSPACE COMPONENTS Vertical vacuum furnaces acuum furnaces are used for hot zone. A metallic hot zone also ex- various thermal treatments in- periences double the heat loss of a are especially suited to heat cluding brazing, , so- graphite hot zone, thus requiring more treat parts having a V lution treatment, aging, hard- energy use for a given cycle. ening, and stress relieving, each of Heating tests were conducted in an cylindrical cross section or which can require different tempera- Ipsen Global V 6060 vertical vacuum parts that can be stacked ture and vacuum ranges, cooling ca- furnace (Figure 1) having an all pability, and enhanced temperature graphite hot-zone construction with vertically, such as those uniformity. Hot zone construction can 50-mm thick insulation and an in- commonly processed in the be either all graphite or all metallic ternal cooling system to evaluate the depending on the intended use and performance of convection versus radi- aerospace industry heating can be via radiation or con- ation heating, temperature uniformity, (e.g., jet engine and turbine vection or a combination. In some and heat loss. Convection-assisted processes, especially in the aerospace heating reduces workload heat up components). Vertical industry, certain parts cannot be heat cycle time, the degree of cycle time re- vacuum furnaces offer high treated in the presence of graphite. For duction being dependent on part example, some titanium parts are con- geometry and process cycle parame- reliability, a smaller installed taminated by graphite and some ters. Cycle time savings of up to 30% footprint, and ease of brazing applications will not work in for higher temperature processes and the presence of graphite. In these cases, 50% for lower temperature processes maintenance. an all-metallic hot zone is required. are possible. Convection heating typ- Metallic hot zones offer a pure envi- ically is used at 2-bar furnace pressure Craig Moller ronment for these special applications. for optimum cycle time savings and is Ipsen International Inc. They also allow the furnace to achieve effective only in furnaces having hot Cherry Valley, Ill. lower vacuum levels at a faster rate zones of graphite construction. and hold a lower vacuum level during the process. However, metallic hot Convection versus Radiation Heating zones are more expensive to purchase Heating test were conducted using a and maintain, and the life expectancy load consisting of four stainless is only about one half that of a graphite cylinders having 12-mm (~0.5 in.) thick

Figure 1 — Ipsen International Global V 6060 bottom-loading vertical vacuum furnace.

Figure 2 — Stainless steel cylinders with 12-mm thick walls used for heating tests; load plus grid support weighs 1,200 kg. HEAT TREATING PROGRESS • MAY/JUNE 2006 25 Radiation to 650oC/Radiation to 1100oC walls set onto a stainless steel heating time was only 110 minutes (a 1200 grid. The total weight of work- 27% time reduction). Fastest part 1000 Slowest part load and grid was about 1,200 kg Average (2640 lb). were Temperature Uniformity Testing C o 800 placed in 12 mm diameter slugs, Five trimmable banks of heating el- which were welded to the out- ements in the all-graphite hot zone are 600 side of the cylinders at the top, located at the top and bottom ends, as well as three banks in the cylindrical 400 middle, and bottom at various lo- cations. Four thermocouples portion of the hot zone arranged as 200 were located at the top, four at top, middle, and bottom (Figure 4). A the bottom, and one in the fixture containing nine Type K thermo- 10 30 50 70 90 110 130 150 middle (Figure 2). The load was couples was used to test temperature Time, min heated to a temperature of 650°C uniformity; four thermocouples were Convection to 650oC/Radiation to 1100oC 1200 (1200°F) at 22°C/min (40°F/min) located on top, four on the bottom, and Fastest part ramp rate using both convection one in the middle (Figure 5). 1000 Slowest part Radiation heating was tested at three Average (2 bar) and radiation heating, temperatures of 540, 815, and 1095°C

C Temperature, after which it was heated from

o 800 650 to 1100°C (1200 to 2010°F) at (1000, 1500, and 2000°F). Convection 600 22°C/min ramp rate using radia- heating was tested at temperatures 540 tion (Figure 3). 400 Water cooled motor Temperature, Temperature, The graphs illustrate that con-

200 vection heating to 650°C is faster and more uniform than radiation Heat heating. During radiation heating exchanger 10 30 50 70 90 110 Time, min to 650°C, the maximum tempera- Figure 3 — Heat up rates in a GV 6060 ver- ture spread was 400°C (720°F), tical vacuum furnace for radiation heating (top) Plenumless while during convection heating to hot zone and convection plus radiation heating (bottom) 650°C, the maximum spread was only for a 1,200 kg load. 200°C (360°F). Radiation heating to Heating 650°C required 100 minutes, while con- elements Gas vection heating required only 60 min- nozzle utes, a 40% time reduction. The overall cycle time to 1100°C was also shorter using the convection/radiation com- Convection bination. Radiation heating for the fan complete cycle required 150 minutes Conventional Bottom pack heating time, whereas with the con- fan motor gas nozzles vection/radiation combination, Figure 4 — Schematic cross section of Global V 6060 vertical vacuum furnace showing com- ponents and internals.

Insulation hot face

Heating position

Cooling position

Figure 6 — Cooling gas nozzle with blocking flaps prevents intrusion of cold gas into the hot Figure 5 — Temperature uniformity test fixture with attached thermocouples. zone when using convection heating. 26 HEAT TREATING PROGRESS • MAY/JUNE 2006 and 815°C). Temperature uniformity is good (4°C), but the 'T = 6.1oC 539.9oC control for convection heating was tested with readings were all below setpoint. 560 cooling nozzles closed and open to ob- Figure 8 shows the temperature serve the effect of blocking flaps (the spread for the convection heating uni- 550 C patented cooling nozzle design shown formity test is good at both 540 and o in Figure 6 incorporates blocking flaps 815°C (6 and 8°C, respectively) and is 540 to prevent the intrusion of cold gas into fairly well centered about the control the hot zone when using convection point. Further improvement of tem- Temperature, 530 heating). During the uniformity perature uniformity could be achieved testing, the five heating zones were ad- by incorporating independent zones 520 justed only once for best uniformity at of control in the furnace, which would 0 1 2 3 4 5 6 7 8 9 10 Thermocouple location 540°C, and those settings were used allow for each heating zone to inde- 'T = 8.3oC 815.6oC control for the other temperatures. The heating pendently adjust uniformity at any set 835 zones required different adjustments temperature. Data from the nozzle at 540°C for radiation and convection flaps closed and held open (Figure 9)

C 825 heating. during convection heating shows that o Figure 7 shows that radiation uni- uniformity is not good without the 815 formity at 540 and 815°C is very good, blocking effect of the flaps; the temper- and the temperature spread at 1095°C ature spread is about double when the Temperature, 805 flaps are open. This test confirms the 'T = 2.8oC 538.9oC control 560 benefit of the cooling nozzle design. 795 0 1 2 3 4 5 6 7 8 9 10 Thermocouple location 550 Heat Loss Test Results C o Steady-state power measurements Figure 8 — Convection heating uniformity in GV 6060 vacuum furnace with graphite hot were taken at each soak temperature 540 zone at test temperatures of 540°C (top) and during the heating uniformity tests to 815°C (bottom).

Temperature, Temperature, determine the amount of heat lost 16 530 through the insulation and into the 14 Nozzles C o water jacket of the vacuum furnace. 12 Closed 520 Open 0 1 2 3 4 5 6 7 8 9 10 These heat loss tests were conducted 10 Thermocouple location for both radiation and convection 8 'T = 3.3oC 815.6oC control heating. Heat loss was measured in 6 835 both cases with the cooling nozzle 4

flaps closed and held open. Figure 10 spread, Temperature 2 825 shows that heat loss for convection 0

C 538 816 o heating benefits from the blocking ef- Soak temperature, oC

815 fect of the nozzle flaps. Heat loss is 60 Figure 9 — Convection heating uniformity to 85% greater when the flaps were with nozzle flaps open and closed.

Temperature, Temperature, held open. For radiation heating, there 250 805 was an insignificant change in the heat kW - closed loss with flaps open versus closed. 200 kW - open 795 150 0 1 2 3 4 5 6 7 8 9 10 Thermocouple location Furnace Cooling Performance The same test load used for the 100 'T = 3.9oC 1092.8oC control Heat loss, kW 1110 heating tests was used to test the 50 cooling capacity of the furnace. The 0 cooling system is designed so the hot 538 816 o 1100 gas from the load enters the cooling Soak temperature, C C o fan first and is then is forced through 120 kW - closed 100 1090 the heat exchanger (refer to Figure 4). kW - open This arrangement ensures complete 80

Temperature, Temperature, coverage of the heat exchanger by the 60 1080 wind flow coming off the fan. It also 40

permits the load in the furnace to be Heat loss, kW 20 1070 cooled to a temperature near that of 0 1 2 3 4 5 6 7 8 9 10 the incoming water, as there is no fur- 0 Thermocouple location 538 816 1093 ther compression of the cooling gas o Figure 7 — Radiation heating uniformity in Soak temperature, C GV 6060 vacuum furnace with graphite hot after it leaves the heat exchanger. The Figure 10 — Steady-state heat loss at soak zone at test temperatures of 540°C (top), 815°C nominal rating of the water-cooled temperature for convection (top) and radiation (middle), and 1095°C (bottom). quench motor is 160 kW. Water cooling (bottom) heating. HEAT TREATING PROGRESS • MAY/JUNE 2006 27 1200 using nitrogen cooling gas. Increasing the size of the cooling fan would allow Fastest part using the extra power capacity of the 1000 Slowest part water cooled motor. The cooling rate Average achieved in this test also demonstrates C o 800 that the special cooling gas nozzle flaps function correctly by opening to allow 600 the cooling gas to flow into the work-

Temperature, Temperature, zone. 400 An arrangement of bottom cooling gas injection nozzles at the bottom 200 end of the hot zone insulation pack and cooling nozzles around the cylin-

0 drical portion of the hot zone provides 0 5 10 15 20 25 30 faster cooling of large massive parts Time, min where the majority of mass is near the Figure 11 — Cooling rate test results using removes heat from the winding sec- hearth. This is typical of parts that will 2-bar nitrogen to cool 1,200 kg stainless steel tion of the motor, which allows the be clamped in a fixture. The cooling cylinders and support grid workload (refer to motor to develop up to 250% of its gas exits at the top end of the Figure 2) in GV 6060 vertical vacuum furnace. nominal rated power for short periods hot zone. of time, resulting in faster cooling of parts during the high temperature crit- ical range. For more information: Craig Moller, me- The cooling test was run in two-bar chanical engineering manager, Ipsen Inter- nitrogen gas. Test results shown in national Inc., PO Box 6266, Rockford, IL Figure 11 indicate that the cooling rate 61125-1266; tel: 815-332-2635; fax: 815-332- of the test load is similar to those of 4995; e-mail: [email protected]; Internet: other two-bar furnaces www.ipsen-intl.com.

28 HEAT TREATING PROGRESS • MAY/JUNE 2006