You Can’t Buy a Vacuum Furnace This Year. You Can still buy capacity! CFC Fixturing
(Editors Note; The team at “The Monty” has seen first hand how replacing cast fixturing with lighter weight CFC fixturing in vacuum heat treating applications has resulted in increased throughput representing an ROI on CFC fixturing of less than 12 months-read on).
“Turbine demand is running past furnace supply, and a new vacuum furnace is still a year or two out. Some of the capacity you need is already sitting inside the furnaces you own — tied up in the fixture.
Last July, “The Monty Heat Treat News” reported that the AI data center buildout is pushing gas turbine demand into territory nobody planned for, and that vacuum furnace orders are following right behind it. Turbine lead times past five years. Siemens Energy nearly doubling unit sales in a single year. Mitsubishi planning to double manufacturing capacity by 2027.
Here is where that lands on the shop floor. A new vacuum furnace is itself 12 to 24 months out. So the heat treater who has been told to push 30 or 40 percent more through the department this year is not going to get there with a purchase order. He gets there, or doesn’t, with the furnaces already bolted to his floor.
Which makes one question worth asking out loud: how much of your existing furnace capacity is currently occupied by the fixture?
Inside that cycle the fixture is doing three jobs nobody wrote down. It is a heat sink, so it sets how long the load takes to equalize before the filler melts. It is an obstruction, so it shapes the gas flow during quench, and quench rate drives gamma prime size directly. And it is a mechanical boundary, holding thin-wall parts that will creep under their own weight at 1,200°C if the support points are wrong.
Here is the failure mode that frustrates process engineers most. The furnace passes its temperature uniformity survey. The recipe is unchanged and still approved. The operator did nothing differently. And braze quality drifts anyway, because the metal fixture has crept two or three millimeters over eighty cycles, the joint gap has shifted, and the support points are no longer where the qualification assumed they were. The evidence shows up in metallography weeks later, on parts that have already absorbed casting, coating and cooling hole drilling. You are not scrapping raw stock. You are scrapping finished airfoils.
What carbon fixtures actually do, and what they don’t; Carbon fiber reinforced carbon has an obvious appeal here, and an equally obvious sales pitch attached to it that deserves to be retired. Five times lighter does not mean five times less heat. The specific heat capacity of carbon runs around 1.8 to 2.0 kJ/(kg·K) against roughly 0.65 to 0.75 for nickel-base alloys. The honest net benefit is on the order of 1.5 to 2. That number is smaller than the brochure version and it survives being checked, which is the only kind of number worth putting in front of a process engineer.
The carburization question is the one that actually decides whether CFC gets through the door, and it has two separate answers because there are two separate paths. Direct contact above about 1,100°C will form chromium carbide in the surface zone of a nickel-base part. That path is closed by putting spacers of lanthanum oxide doped molybdenum between the carbon and the component. Molybdenum holds its shape at process temperature, it is not wetted by BNi filler metals so stray braze does not weld the part to its support point, and the lanthanum oxide dispersion keeps the material ductile even after recrystallization, where unalloyed molybdenum turns brittle. Worth saying plainly: those spacers sit in carbon contact themselves and form molybdenum carbide over cycle count. They are consumables with a service life, not fit-and-forget hardware.
The second path is the gas phase, CO forming from residual moisture and oxygen at the carbon. No contact barrier touches that one. It is handled by bake-out and preconditioning before the first production run, and any supplier who tells you a spacer eliminates carburization outright has skipped half the mechanism.
Two more things have changed on the fixture side that are worth knowing about. Load peaks can now be calculated at process temperature and designed out of the structure, instead of being buried under a safety allowance of extra mass — and since mass is heat sink, that feeds straight back into equalization time. The calculation is steady state, so it covers sag and support points at the hold, not the heat-up and quench transients. And fixtures can be individually marked, which sounds trivial until you connect it to the failure mode above. Mark the fixture and its position in the load, and cycle count, position and history become traceable. The variable that was invisible becomes a line in the record.
The part the purchasing side should read; Nobody in this market is being measured on fixture savings. On price you can win 20 percent. On schedule you can lose everything, because turbine contracts carry penalties and a failed load with the rotor shop waiting costs a multiple of the annual fixture budget. Meanwhile furnaces are sold out. Any fixture change that raises parts per load or shortens cycle time converts directly into capacity, at a fraction of furnace price and none of the furnace lead time. How much capacity is a question each shop has to answer with its own part family and its own numbers. The point is that this year it is the only capacity available. And modular systems beat discounts, because a separate drawing number for every part family costs more in administration than any rebate returns.
How to find out for yourself; Pick one part family with high volume and a known distortion problem. Record what you have now: furnace type, usable envelope, recipe, current fixture mass, parts per load, cycle time, fixture service life, documented scrap causes. Agree acceptance criteria in writing before anything goes in the furnace — part dimensional stability, metallography of the surface zone for carbide formation, joint gap, cycle time, parts per load. Then run a defined number of cycles and measure the fixture before and after. That measurement is the whole argument. Everything else is a claim.
Your furnace is qualified. Your recipe is qualified. Your fixture changes every eighty cycles and appears in none of your records.”
Related Articles From “The Monty Heat Treat News”;
“CFC Fixturing-A Growing Trend?”
“Gear Supplier Invests in CFC Fixturing”
“CFC Fixtures Even For Extreme Conditions”
