Anyone who has spent time on a metrology lab floor knows that the base under a coordinate measuring machine matters just as much as the sensor head on top of it. A probe rated to sub-micron repeatability is only as good as the surface it sits on. Yet base material selection is often treated as an afterthought, decided by habit or by whatever the last supplier offered, rather than by the actual physics of the application.
There are really three candidates in play for most precision platforms: natural granite, cast iron, and welded or machined steel. Each has a legitimate place in industrial equipment. The differences show up not in a spec sheet's headline numbers, but in how each material behaves once it leaves the factory and starts living in a real workshop.
Thermal Behavior Is Where the Differences Start
Steel and cast iron are metals, and metals move. A steel weldment with a thermal expansion coefficient around 11–12 µm/m·°C will visibly drift over the course of a single day in a shop where the temperature swings even a few degrees between morning and afternoon. For woodworking or general machine tool frames, that drift is irrelevant. For a CMM holding tenths-of-a-micron tolerances, it is the difference between a passing part and a rejected one.
Granite's thermal expansion coefficient runs roughly 5–8 µm/m·°C depending on the specific quarry and mineral composition, and more importantly, granite has low thermal conductivity. It absorbs and releases heat slowly, which means it doesn't "flinch" every time a door opens or a machine two meters away kicks on. In climate-controlled rooms this matters less, but very few production floors are perfectly climate-controlled around the clock, and even in rooms that are, transient heat sources (operators, lighting, adjacent equipment) still create local gradients that a metal base will register almost immediately.
Damping and Vibration
Cast iron has a reputation for good vibration damping, and it's earned - the graphite flakes in its microstructure do dissipate energy effectively, which is why it remains the standard choice for many machine tool beds. But natural granite, being a naturally consolidated crystalline structure rather than a cast or welded one, typically shows internal damping capacity several times higher than cast iron in comparable geometries. That's part of why granite surface plates became the default reference standard in metrology labs starting in the mid-20th century - it wasn't nostalgia, it was that a granite plate settles down faster after being disturbed and stays quieter while measurements are taken.
Steel, especially welded steel, is the weakest performer here unless it's heavily ribbed or filled with damping compound, because welded joints introduce their own resonances and residual stress that can very slowly relax over months or years - exactly the kind of slow dimensional creep you don't want under a reference instrument.
Long-Term Dimensional Stability
This is the one that surprises people who haven't worked with all three materials over a multi-year equipment lifecycle. Cast iron and steel are both subject to gradual relaxation of internal stresses from casting or welding. Manufacturers manage this with stress-relief annealing and aging processes, and a well-made cast iron base can be very stable - but "well-made" requires a slow, deliberate cooling and aging cycle that many lower-cost suppliers skip or shorten to save time.
Granite doesn't have this problem in the same way, because it was formed geologically over millions of years and doesn't carry the same kind of internal casting stress. What it does need is correct handling: proper support points during grinding, controlled humidity during lapping, and a supplier who actually understands where a slab's natural grain runs. A granite plate ground without attention to grain direction can still warp slightly over time, just for different metallurgical reasons than a metal one.
Cost, Weight, and Practical Trade-offs
None of this means granite wins every application. Granite is brittle - it doesn't like shock loading or being dropped, and it can't be tapped, drilled, or welded on-site the way steel can. For moving machine frames, gantries, or anything that needs threaded mounting points added after the fact, steel or cast iron is usually still the right call. Granite is also denser per unit stiffness in some geometries, which can be an advantage (mass helps damping) or a disadvantage (harder to move, higher shipping cost) depending on the installation.
The practical rule most metrology engineers use: if the equipment measures rather than moves, and if it needs to hold a reference tolerance over years rather than months, granite is usually worth the extra procurement effort. If the equipment cuts, presses, or needs field modification, cast iron or steel remains the more sensible choice.
What to Ask a Supplier
Whichever material ends up on the quote, a few questions tend to separate a serious supplier from one that isn't: What is the actual density of the granite being used, and can they provide a mill or quarry test certificate? Marble is sometimes substituted for granite by less scrupulous fabricators because it's cheaper and superficially similar - but marble is softer, more porous, and dimensionally less stable, so density and hardness verification matters. For metal bases, ask about the stress-relief process and how long the casting was aged before final machining. And for any base material, ask what calibration standard was used to verify flatness after finishing - a DIN 876, ASME B89.3.7, or JIS B7513 grade should be stated explicitly, not just implied.
At UNPARALLELED, our granite base components are machined and lapped in-house using instruments traceable to national metrology institutes, and every plate ships with its calibration certificate. But regardless of who supplies the base, the underlying physics doesn't change - and understanding it is the first step to specifying equipment that still holds tolerance five years from now, not just on the day it's installed.