A metrology lab manager once called us after a CMM recalibration came back with results that didn't match the machine's rated accuracy. The machine itself was fine. The problem was underneath it - a "granite" base that turned out to be dyed marble, sourced from a supplier who quoted 15% below everyone else. Six months of thermal cycling in the shop had let it creep out of flatness by several microns.
This kind of substitution happens more often than most buyers realize, and it's worth understanding why the difference matters before it costs you a recalibration cycle - or a batch of scrapped parts.
Marble and granite are not interchangeable materials
Both are natural stone, both can be polished to a mirror finish, and to someone without a materials background, a slab of each can look nearly identical once cut and lapped. But their internal structure is completely different. Marble is a metamorphic rock composed mainly of calcite, which is chemically reactive and relatively soft - that's part of why it's prized for sculpture and countertops. Granite is an igneous rock, formed from slowly cooled magma, dominated by quartz and feldspar crystals locked into a dense interlocking matrix.
That crystalline structure is what gives granite its dimensional stability. Black granite typically used in precision applications runs around 3,100 kg/m³ in density, noticeably denser than most commercial marble. Higher density generally correlates with lower porosity, which means less moisture absorption, less thermal expansion variance, and less long-term creep under load - exactly the properties you need in something that's supposed to hold a reference plane for years without drifting.
Marble, by contrast, has a coarser and more variable crystal structure. It's more prone to microscopic fissures, absorbs moisture more readily, and its coefficient of thermal expansion tends to be less uniform across the slab. On a kitchen countertop none of this matters. Under a coordinate measuring machine that's expected to hold submicron repeatability, it matters a great deal.
Why some suppliers substitute marble anyway
The honest answer is cost and processing speed. Marble is generally softer and faster to grind and polish, which shortens production time and lowers tooling wear. A base that looks finished and flat off the line can pass a quick visual or even a basic straightedge check. The problems show up later - after months of temperature cycling in a shop environment, after repeated loading and unloading of workpieces, after the stone has had time to relax into whatever internal stresses were locked in during quarrying and cutting.
This is a genuine issue in the industry, not a hypothetical one. Buyers sourcing on price alone, especially for large single-piece bases, sometimes don't find out what they actually received until a precision audit flags an out-of-spec result. At that point the cost of the "savings" has usually been eaten several times over in downtime and requalification.
What to actually check before you buy
If you're specifying a granite base and want to verify what you're getting, a few checks go a long way:
Density. Genuine precision-grade black granite should sit around 3,000–3,100 kg/m³. A supplier who can't or won't provide a density figure, or who gives you a number well below that range, is a red flag worth pursuing further.
Hardness and scratch resistance. Granite is dramatically harder than marble - a simple scratch test with a steel point isn't a lab-grade method, but a stone that scratches easily under light pressure is not precision granite.
Calibration traceability. Ask for the calibration certificate on the finished surface plate or base, and check which standards body issued it. Reputable measurement labs calibrate to national metrology institute traceability, not just an in-house gauge block comparison.
Standard reference. Flatness and grade tolerances for granite surface plates are defined in standards like DIN 876, ASME B89.3.7, and JIS B7513, among others. If a quote references "precision grade" without pointing to a specific standard and grade, ask which one.
Thermal soak history. Ask how long the raw block was aged and stress-relieved before final grinding. Stone that's rushed from quarry to finished product hasn't had time to release internal stress, and it tends to move after installation regardless of how flat it measured on day one.
The stakes get higher as tolerances tighten
None of this matters much for a workbench or a decorative surface. It matters enormously for anything sitting under a CMM, a laser interferometer, an AOI system, or a wafer-handling stage, where the base isn't just structural - it's part of the measurement chain. A base that moves by even a few microns over a year of thermal cycling introduces an error that no amount of software compensation fully corrects, because the error isn't constant; it drifts with temperature, humidity, and load history.
For equipment builders and metrology labs, the practical takeaway isn't "always distrust your supplier" - it's "always ask for the numbers." Density, hardness, calibration traceability, and stress-relief history are all things a legitimate precision granite manufacturer should hand over without hesitation. If those answers are vague, that vagueness is itself the data point worth paying attention to.