{"id":13341,"date":"2026-10-08T07:16:36","date_gmt":"2026-10-08T07:16:36","guid":{"rendered":"https:\/\/www.vacuum-sintering.com\/vacuum-brazing-furnace\/"},"modified":"2026-10-08T07:16:42","modified_gmt":"2026-10-08T07:16:42","slug":"four-de-brasage-sous-vide","status":"publish","type":"post","link":"https:\/\/www.vacuum-sintering.com\/fr\/vacuum-brazing-furnace\/","title":{"rendered":"Four de brasage sous vide\u00a0: \u00e0 quoi servent r\u00e9ellement une pression de 6,7 \u00d7 10\u207b\u00b3 Pa et une zone chaude en molybd\u00e8ne \u00e0 1\u00a0300\u00a0\u00b0C\u00a0?"},"content":{"rendered":"

A vacuum brazing furnace is a batch furnace that joins an assembly by melting a filler metal inside a controlled high-vacuum or hydrogen atmosphere \u2014 in our H series, 6.7 \u00d7 10\u207b\u00b3 Pa with a molybdenum (MO) hot zone rated 1000 \/ 1300 \u00b0C \u2014 so that a component carrying dozens of separate joints is bonded in a single thermal cycle, without flux, without a melted parent metal, and without a heat-affected zone.<\/strong> Welding creates a joint by melting the parts themselves; brazing creates one by melting something else, and that single distinction is why the furnace \u2014 not the torch \u2014 decides whether the joint is continuous.<\/p>\n

Most teams that go looking for a four de brasage sous vide<\/em> are not shopping for heat. They already know the temperature. What they need is a machine that can hold a residual oxygen partial pressure low enough that a molten filler will wet a real production surface, on parts that have been handled, stacked, fixtured and possibly stored for weeks.<\/p>\n

This article covers what the vessel has to deliver, the six stages of a cycle, and the parameters worth writing into a purchase specification.<\/p>\n

\"vacuum-brazing-furnace-horizontal-molybdenum-hot-zone<\/figure>\n

Qu'est-ce qu'un four de brasage sous vide\u00a0?<\/h2>\n

A brazing furnace has to deliver four things at the same time. Any one of them missing turns a good filler into a bad joint:<\/p>\n\n\n\n\n\n\n\n\n
Requirement<\/th>\nWhy it decides the outcome<\/th>\nH series specification<\/th>\n<\/tr>\n<\/thead>\n
High vacuum<\/strong><\/td>\nWetting depends on the oxide state of the surface, and oxide state depends on the oxygen partial pressure above the part \u2014 not on “how clean the part looked”<\/td>\n6.7 \u00d7 10\u207b\u00b3 Pa (diffusion \/ molecular pump class)<\/td>\n<\/tr>\n
An isothermal hot zone<\/strong><\/td>\nCapillary flow needs the whole joint at the same temperature; a hot spot drains the filler away from the cold end<\/td>\nMolybdenum (MO) heating element, 1000 \/ 1300 \u00b0C<\/td>\n<\/tr>\n
A hydrogen option<\/strong><\/td>\nVacuum prevents new oxide from forming. A reducing atmosphere can remove oxide that is already there \u2014 a different job<\/td>\nVacuum and hydrogen integrated in one vessel<\/td>\n<\/tr>\n
Enough usable volume<\/strong><\/td>\nBrazing economics come from amortising one pump-down and one ramp across many joints<\/td>\n400 \u00d7 400 \u00d7 600 mm up to 800 \u00d7 800 \u00d7 1200 mm; vertical \u03a6650 \u00d7 1200 mm and \u03a6650 \u00d7 3200 mm<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

Two clarifications worth making before a specification is written.<\/p>\n

First, the vacuum figure is a pump-group configuration class<\/strong>, not a precision vanity number. 6.7 \u00d7 10\u207b\u00b3 Pa is the high-vacuum class reached with a diffusion or molecular pump. A furnace quoted at 1 Pa or 10 Pa is running on mechanical pumping only and is intended for a different duty. If your process logic needs genuine high vacuum at the joint, say so at enquiry stage, because that decision fixes the whole machine.<\/p>\n

Second, the chamber number that matters is the usable<\/strong> volume \u2014 the volume left after the hot zone, the fixture and the loading table are in place. A 800 \u00d7 800 \u00d7 1200 mm vessel does not accept an 800 \u00d7 800 \u00d7 1200 mm assembly.<\/p>\n

How Vacuum Brazing Works \u2014 Stage by Stage<\/h2>\n

Stage 1 \u2014 Cleaning, gap control and assembly<\/h3>\n

Vacuum brazing has no flux, and that is both the advantage and the discipline. Nothing in the process will dissolve a fingerprint, a cutting-fluid residue or an oxide film. Plate-fin and tube-and-header assemblies are typically tacked, pinned or fixture-clamped at a joint clearance in the 0.02\u20130.10 mm band \u2014 tight enough for capillary force to pull the filler through, wide enough for the filler to actually flow.<\/p>\n

Stage 2 \u2014 Pump-down and outgassing<\/h3>\n

The pump-down is not a waiting period; it is a process step. Water vapour and hydrocarbon residues desorb from the part, the fixture and the vessel walls, and most of them leave in the 200\u2013400 \u00b0C window. Rushing through that window traps vapour at the joint and shows up later as voids along the bond line. A holding segment during heat-up is therefore standard practice.<\/p>\n

Stage 3 \u2014 Heating to the filler’s flow temperature<\/h3>\n

Our H series ramps at 1\u201315 \u00b0C\/min up to 1300 \u00b0C. That range is a decision, not a limitation. A fast ramp on a steel-to-steel assembly is often harmless. The same ramp across a ceramic-to-metal assembly \u2014 alumina or aluminium nitride sealed to copper, Kovar or a nickel-iron alloy \u2014 will crack the ceramic, because the differential expansion is resolved by stress rather than by time. Above the filler liquidus, the soak is typically held 20\u201350 \u00b0C above it, long enough for capillary flow and short enough to limit interdiffusion between the filler and the base metal.<\/p>\n

Stage 4 \u2014 Soak: melt, wet, flow<\/h3>\n

This is the stage people picture, and it is the shortest. The hold is typically 5\u201330 min at temperature for a filler that flows well; nickel-based fillers on stainless steel and superalloys need the longer end, silver-based fillers on copper and brass the shorter. The observable end point is not a timer \u2014 it is a filler fillet that has formed on both sides of the joint.<\/p>\n

\u00c9tape 5 \u2014 Refroidissement contr\u00f4l\u00e9<\/h3>\n

Cooling is the stage that gets designed last and causes the most rework. The filler must solidify while both members are still dimensionally stable. Where the joint is between dissimilar materials, the cooling ramp is set by the member with the highest coefficient of thermal expansion, not by the largest mass. Forced cooling is deliberately avoided at this point in a brazing cycle for exactly this reason.<\/p>\n

What changes when you switch to hydrogen<\/h3>\n

Our H series runs vacuum and hydrogen in the same vessel, and the switch is not cosmetic. Vacuum protects passively \u2014 no oxygen, so no new oxide forms. Hydrogen protects actively: copper, nickel and iron oxides are chemically reduced, which is why stainless-steel and copper assemblies can be brazed flux-free. The controlling variable is the moisture budget \u2014 the dew point of the gas reaching the work \u2014 and that is a measurement and control problem, not a gas-supply problem. Hydrogen is also unforgiving to alloys containing high-vapour-pressure elements, and it requires purge sequencing, exhaust handling and interlock hardware engineered into the machine rather than added later.<\/p>\n

Key Parameters \u2014 What to Put in the Specification<\/h2>\n\n\n\n\n\n\n\n\n\n\n\n
Param\u00e8tre<\/th>\nOur H series<\/th>\nWhat it actually controls<\/th>\nQuestion worth asking a supplier<\/th>\n<\/tr>\n<\/thead>\n
Aspirateur ultime<\/td>\n6,7 \u00d7 10\u207b\u00b3 Pa<\/td>\nOxygen partial pressure at the joint; filler wetting<\/td>\nWhich pump set is included, and what leak rate is guaranteed at that level?<\/td>\n<\/tr>\n
Temp\u00e9rature maximale<\/td>\n1000 \/ 1300 \u00b0C<\/td>\nWhich filler classes are addressable<\/td>\nIs the rating for the hot zone or for the work zone?<\/td>\n<\/tr>\n
vitesse de chauffage<\/td>\n1\u201315 \u00b0C\/min (to 1300 \u00b0C)<\/td>\nThermal gradients in dissimilar joints<\/td>\nCan the profile differ between the vacuum stage and the hydrogen stage?<\/td>\n<\/tr>\n
Atmosph\u00e8re<\/td>\nVacuum + hydrogen, one vessel<\/td>\nPassive vs. active oxide control<\/td>\nHow is dew point monitored and recorded during the hydrogen stage?<\/td>\n<\/tr>\n
Chamber (horizontal)<\/td>\n400 \u00d7 400 \u00d7 600 \/ 600 \u00d7 600 \u00d7 900 \/ 700 \u00d7 700 \u00d7 1000 \/ 800 \u00d7 800 \u00d7 1200 mm<\/td>\nLoad size and stacking height<\/td>\nWhat is the usable volume inside the hot zone with fixture fitted?<\/td>\n<\/tr>\n
Chamber (vertical)<\/td>\n\u03a6650 \u00d7 1200 \/ \u03a6650 \u00d7 3200 mm<\/td>\nLong, shaft-like and tubular assemblies<\/td>\nHow is the load supported, and does the support conduct heat away?<\/td>\n<\/tr>\n
Loading direction<\/td>\nSide (horizontal models) \/ bottom, top (vertical models)<\/td>\nFixturing strategy and gravity-induced distortion<\/td>\nWhich surface are the critical joints resting on?<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

Models We Build \u2014 H Series Vacuum Hydrogen Furnace<\/h2>\n\n\n\n\n\n\n\n\n\n\n
Mod\u00e8le<\/th>\nStructure<\/th>\nChambre (mm)<\/th>\nHeater<\/th>\nChargement<\/th>\nUltimate vacuum (Pa)<\/th>\nMax temperature (\u00b0C)<\/th>\n<\/tr>\n<\/thead>\n
HV6S\/120<\/td>\nVerticale<\/td>\n\u03a6650 \u00d7 1200<\/td>\nMO<\/td>\nBottom<\/td>\n6.7 \u00d7 10\u207b\u00b3<\/td>\n1000 \/ 1300<\/td>\n<\/tr>\n
HV6S\/320<\/td>\nVerticale<\/td>\n\u03a6650 \u00d7 3200<\/td>\nMO<\/td>\nTop<\/td>\n6.7 \u00d7 10\u207b\u00b3<\/td>\n1000 \/ 1300<\/td>\n<\/tr>\n
H4M013<\/td>\nHorizontal<\/td>\n400 \u00d7 400 \u00d7 600<\/td>\nMO<\/td>\nSide<\/td>\n6.7 \u00d7 10\u207b\u00b3<\/td>\n1000 \/ 1300<\/td>\n<\/tr>\n
H6M013<\/td>\nHorizontal<\/td>\n600 \u00d7 600 \u00d7 900<\/td>\nMO<\/td>\nSide<\/td>\n6.7 \u00d7 10\u207b\u00b3<\/td>\n1000 \/ 1300<\/td>\n<\/tr>\n
H7M013<\/td>\nHorizontal<\/td>\n700 \u00d7 700 \u00d7 1000<\/td>\nMO<\/td>\nSide<\/td>\n6.7 \u00d7 10\u207b\u00b3<\/td>\n1000 \/ 1300<\/td>\n<\/tr>\n
H8M013<\/td>\nHorizontal<\/td>\n800 \u00d7 800 \u00d7 1200<\/td>\nMO<\/td>\nSide<\/td>\n6.7 \u00d7 10\u207b\u00b3<\/td>\n1000 \/ 1300<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

How to read this table: the horizontal, side-loading models<\/strong> suit flat stacks, plate-fin cores and manifold assemblies, where the load is wide and shallow. The vertical models<\/strong> exist for parts long relative to their section \u2014 tubular assemblies, shafts and sealing components \u2014 up to 3200 mm. Choosing between the two is usually a fixturing decision before it is a capacity decision.<\/p>\n

Vacuum Brazing vs. Other Joining Routes<\/h2>\n\n\n\n\n\n\n\n\n\n
Itin\u00e9raire<\/th>\nHow the joint forms<\/th>\nTypical process window<\/th>\nDistortion \/ HAZ<\/th>\nGeometry freedom<\/th>\nWhere it wins<\/th>\n<\/tr>\n<\/thead>\n
brasage sous vide<\/strong> (H series)<\/td>\nFiller melts and is drawn into the gap by capillary action<\/td>\nUp to 1300 \u00b0C; 6.7 \u00d7 10\u207b\u00b3 Pa<\/td>\nVery low; entire load heated uniformly, non-melted base metal<\/td>\nAny geometry that fits the chamber; many joints per cycle<\/td>\nMulti-joint assemblies, thin-wall and plate-fin structures, ceramic-to-metal<\/td>\n<\/tr>\n
Hydrogen-atmosphere brazing<\/strong> (same H series)<\/td>\nSame mechanism, active reducing atmosphere<\/td>\nSame window; dew point is the control variable<\/td>\nTr\u00e8s bas<\/td>\nSame<\/td>\nCopper, nickel and stainless assemblies needing oxide removal without flux<\/td>\n<\/tr>\n
Liaison par diffusion<\/strong> (D series)<\/td>\nSolid-state; bond forms by atomic diffusion under pressure, with or without an interlayer<\/td>\nUp to 1400 \u00b0C; 6.7 \u00d7 10\u207b\u00b3 Pa; mechanical pressure applied<\/td>\nLow, but pressure and flatness are mandatory<\/td>\nConstrained by press geometry and flat mating faces<\/td>\nDissimilar metals, joints that must match parent-metal properties, no-filler requirements<\/td>\n<\/tr>\n
Hot pressing<\/strong> (P series)<\/td>\nSimultaneous uniaxial pressure and temperature consolidate a powder or preform<\/td>\nUp to 2200 \u00b0C; 100\u20131200 ton; vacuum configuration depends on pump set<\/td>\nUniaxial; density gradients possible<\/td>\nSet by die and ram, not by free geometry<\/td>\nPowders, composites, near-net shapes where pressure must exceed what gas or capillary can deliver<\/td>\n<\/tr>\n
Fusion welding<\/strong> (TIG \/ electron beam)<\/td>\nParent metal is melted locally<\/td>\nLocal melt, no bulk heating<\/td>\nHigher; local heat-affected zone and residual stress<\/td>\nFree-form, unrestricted size<\/td>\nStructural joints on thick sections where a fusion bond is required<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

The decision that catches people out is the middle row: brazing and diffusion bonding are often treated as interchangeable because both are “vacuum joining”. They are not. Diffusion bonding needs pressure and produces a joint without a filler; brazing needs a capillary gap and produces a joint with one. Where the parent metal is itself heat-treated to a specific temper, the brazing temperature \u2014 not the furnace \u2014 sets whether the part survives with its properties intact.<\/p>\n

Applications Where a Batch Brazing Furnace Pays<\/h2>\n