A machine foundation has to carry the load without settling and hold the base rigid or isolated, whichever the OEM spec calls for. Anchor bolts go in cast-in-place, sleeved, or post-installed depending on whether the foundation is new or existing. Grout — epoxy or non-shrink cementitious — fills the gap under the base and must cure to spec before anchors get torqued and leveling can be trusted.
Every conversation about a machine move eventually gets to leveling — center the bubble, check the ways, run a test cut. That conversation assumes something upstream of it: that the machine is sitting on a foundation that will still be flat, still be rigid, and still be exactly where it was poured a year from now. A level check on a foundation that settles, cracks, or transmits vibration it was never built to isolate is a level check on borrowed time. The foundation and the grout under the base are not prep work ahead of the real job — they are the job, for any machine expected to hold tolerance past its first production run.
This is where relocations lose the thread. The rigging gets planned to the inch, the machine survives the move without a scratch, and then it sits on whatever concrete happened to be there, or a new pour designed by someone who never opened the OEM's foundation drawing. What follows is what goes into a foundation built to hold a machine, not just support one.
Does every machine need its own dedicated foundation?
No — it depends on what the machine does to the floor, not just its weight. A light, well-balanced machine tool with a stiff casting can often sit on a properly rated section of an existing slab, set on leveling pads, with no dedicated pour needed. A machine with real dynamic load — a mechanical press with a flywheel and a stroking ram, a forging hammer, a large reciprocating compressor — puts cyclic, impact-type force into whatever it's bolted to, and an existing production slab is rarely built to absorb that. That machine gets its own foundation, sized and reinforced for the load it generates, not for its static weight sitting still.
The OEM's foundation drawing settles this, and it exists for almost anything heavier than a bench-top tool. It specifies pad thickness, reinforcement, bolt pattern, and whether the foundation needs to be isolated from the surrounding floor or tied rigidly into it. Skipping that drawing and pouring a generic slab because "it's just a base" is one of the more expensive guesses in a relocation, since a foundation is nearly impossible to fix once a machine is running on it and the floor around it has been built up to that elevation.
Isolated or rigid — how do you know which mounting a machine actually needs?
By what the machine does to its surroundings, not by what seems safer. A rigid, mass foundation ties the machine solidly into a thick block of concrete, often doweled or keyed into the structural slab below it, so the whole assembly behaves as one stiff body. That's what a grinder, a boring mill, or any machine holding tight geometric tolerance actually wants — a foundation that can flex or rock under cutting load shows up directly as lost accuracy. The added mass also pushes the foundation's natural frequency well clear of the machine's own operating frequencies, so the two don't resonate against each other.
An isolated foundation does the opposite job: it decouples the machine from the building structure on purpose, using spring isolators, elastomeric mounts, or an inertia block set into the floor on a compressible joint, so vibration the machine generates doesn't couple into the slab and reach a CMM or a neighboring process that can't tolerate it. This is the mounting a stamping press, a large compressor, or anything with reciprocating stroke typically needs — and why a metrology room near production traffic gets its own isolated pad, since isolation runs both directions.
Mixing the two up is the actual failure mode. Rigidly grouting a press that needed isolation drives its impact loading straight into the building slab; isolating a grinder that needed rigidity lets the base move under cutting force in exactly the direction its accuracy spec assumes it can't. The OEM spec decides this, and where a machine's dynamic behavior isn't documented, that's a foundation engineering question worth answering before concrete gets poured.
What anchor bolt types actually go under a machine base?
Three, chosen by whether the foundation is new or already there. Cast-in-place anchors — J-bolts or L-bolts — go into a template set before the pour, positioned to match the machine's baseplate hole pattern, and embedded to the depth the anchor manufacturer specifies for the bolt diameter and load. These are the cleanest option when a foundation is built from scratch around a known machine, because the pattern is fixed and correct from the start.
Sleeved anchors solve the problem cast-in bolts can't: they go into the pour inside an oversized sleeve or formed pocket, giving the bolt room to shift laterally before it's locked in. After the machine is set and leveled, the bolt gets nudged into its final position through the baseplate hole and the sleeve gets packed with grout, fixing it where the machine actually needed it rather than where a template guessed weeks earlier. Where the baseplate pattern might not match a drawing precisely, sleeved anchors are what save a re-pour.
Post-installed anchors handle the case cast-in and sleeved anchors can't: an existing slab with no bolt pattern at all. Holes get drilled to spec, cleaned thoroughly, and a chemical epoxy anchor gets injected and set, or a mechanical expansion anchor gets torqued in, depending on the load and whether the pull is static or dynamic. Embedment depth and edge distance come from the anchor manufacturer's data sheet, not a rule of thumb — undersizing either is how a post-installed anchor pulls loose under load it was never tested to hold.
What decides the anchor and grout package
- Whether the foundation is new (cast-in or sleeved anchors) or existing (post-installed, drilled anchors)
- Static weight versus dynamic, cyclic load the machine actually generates in operation
- Isolated or rigid mounting, per the OEM's foundation drawing — never assumed either way
- Oil, coolant, or chemical exposure at the base, which rules out plain cementitious grout over time
- How much schedule the project can give the grout before anchors get torqued and the line runs
Epoxy grout or cementitious grout — which one goes under the base?
Epoxy where oil exposure, dynamic loading, or schedule are tight; non-shrink cementitious where the load is static, the base stays dry, and the project can give it the calendar days to cure the long way. Both are engineered products built for machine bases, not bagged mortar, and both must be non-shrink — a grout that shrinks as it cures leaves voids under the baseplate that a machine's own weight eventually finds, which is the single most common thing that turns a good leveling job into a bad one a year later.
The real difference is chemistry and timeline. Cementitious grout cures through hydration, the same reaction that hardens concrete, so its rated strength is only reached after the full cure window. Epoxy cures through a resin-hardener reaction that doesn't depend on evaporation or humidity, reaching load-bearing strength in roughly a day and full cure within about a week regardless of shop conditions. Epoxy also shrugs off oil, coolant, and hydraulic fluid that soak into the base area of most machine tools over years of service — cementitious grout absorbs and degrades under that exposure, which is why oil-pan and gearbox bases lean epoxy even when the load alone doesn't demand it.
Cementitious grout still wins jobs on cost and placement — it flows and finishes easier at larger pour volumes and doesn't carry epoxy's tighter working-temperature window, which matters on a shop floor that isn't climate controlled. For a static machine on a dry base with an unhurried schedule, it's the standard, proven choice. The decision belongs on the grout manufacturer's data sheet for the specific product, matched against load, exposure, and schedule — not picked because it's what the last job used.
How long does grout actually need to cure before anything happens to it?
Long enough to reach the strength the grout data sheet specifies for the stage in question — that number comes from the grout, not the machine builder's manual, since cure time is a property of the product poured, not the equipment on it. Epoxy typically handles careful loading within a day and reaches full design strength inside about a week; standard non-shrink cementitious grout needs the same multi-week cure structural concrete does, though high-early-strength formulations exist to compress that schedule.
Nothing on site should get rushed past what the grout has actually reached. Anchor bolts get snugged, not torqued, until the grout cures — torquing early stresses a bolt against grout that hasn't developed the strength to resist it. Final leveling and requalification wait for full cure too: a machine leveled against grout still gaining strength will be measurably out of level once it finishes settling. Building the cure window into the schedule up front is what keeps a foundation delay from becoming a surprise the week the machine was supposed to run production.
What's the actual sequence — pour, anchor, grout, level?
Foundation first, anchors set into it, machine set and roughly leveled on temporary supports above the grout gap, grout poured to fill that gap, cure, then final leveling and torque — in that order, and out of order is how the whole stack gets redone. The pad gets poured to the foundation drawing's thickness and reinforcement, with cast-in or sleeved anchors positioned in a template before the concrete goes in, or left open for post-installed anchors if the design calls for drilling later. The machine then lands on leveling wedges, jack screws, or shims resting on the rough foundation surface, kept clear of the concrete by a gap — commonly an inch or two — that becomes the grout bed. That gap gets the machine leveled to the OEM's tolerance before any grout goes near it, because grout fills a gap that's already correct; it doesn't correct one that isn't.
Once the level is set, forms go up around the base perimeter to dam the grout in, with chamfered corners that let trapped air bleed out instead of pocketing under the baseplate. Grout gets poured from one side only and allowed to flow across and fill the space under the base — pouring from multiple sides at once traps air in the middle with nowhere to escape, leaving a void under the machine that nobody finds until the base starts to rock. After the cure window closes, forms come off, anchors get torqued to spec in a cross or star pattern rather than one at a time around the perimeter, and only then does the crew run the final precision level check. That's the handoff into the geometry work covered in machine leveling and alignment — tramming, squareness, and the test cut that proves the machine is back in tolerance, all of it assuming the foundation underneath is done moving.
Bottom line
- Whether a machine needs its own foundation comes down to dynamic load, not static weight — a press or compressor puts cyclic force into the floor that a light machine tool never does.
- Rigid, mass foundations keep precision machines stiff; isolated foundations decouple dynamic equipment from the building. Mixing the two up is the actual failure mode.
- Anchor type follows the foundation: cast-in or sleeved bolts for a new pour, post-installed epoxy or expansion anchors for an existing slab.
- Epoxy grout suits oil exposure, dynamic loads, and tight schedules; non-shrink cementitious suits static, dry, unhurried jobs. Both must be non-shrink.
- Cure time is set by the grout data sheet, not the machine manual — torquing anchors or leveling against uncured grout undoes the accuracy of everything done correctly up to that point.
A foundation gets one chance to be right — fixing it later usually means jacking a running machine back up. Machinery moving lands the machine on its marks over a foundation that's ready for it, and millwright services covers the anchoring, grouting, and leveling that follows, coordinated as one schedule instead of two jobs colliding on move day. Send the foundation drawing, or tell us there isn't one — start here.