3D scanning services in the UK: what accuracy, file formats and cost to expect

Most people searching for a 3D scanning service are really asking one of two questions: can you capture this thing, and what will I be able to do with the result? The first is about the object. The second is about the file you get back, and that is where jobs go wrong. A scan is not a drawing. It is a very detailed cloud of measured points, and whether it is useful depends on what happens next.

This is a buyer's guide to a 3D scanning service: what scans well and badly, what accuracy means in practice, which file formats you can expect, and what moves the price.

What scans well, and what doesn't

Structured-light and laser scanners work by seeing a surface. That makes most solid, matt, mid-sized objects straightforward: cast housings, brackets, trim, tooling, moulded plastic, sculptural detail. Awkward surfaces are the exception, and the usual culprits are shiny metal, clear plastics and very dark or black finishes, because they either bounce the light away or swallow it. Large flat featureless panels can also drift, because the scanner has nothing to lock on to.

None of these is a dead end. A light, removable matt coating deals with most shiny and transparent parts, and positioning markers help with plain or repetitive geometry. What a scanner genuinely cannot do is capture what it cannot see: deep internal bores, undercuts and blind pockets are measured by other means (calipers, bore gauges, a section cut if the part is expendable) and combined with the scan. If your part is mostly internal, say so in the enquiry.

Accuracy: ask what it needs to be, not what the scanner can do

Scanner accuracy is quoted as a headline number, but the accuracy of the finished file is a different thing. It is shaped by the size of the part, the surface, how the scan was aligned, and what is done to the data afterwards. A small, matt part scanned carefully will hold much tighter than a large, shiny one. So the useful question is not how accurate is the scanner, it is how accurate does this part need to be.

For most replacement and fitting jobs, the fit comes down to a handful of critical features: a bore, a mounting face, a mating edge. Those are the ones to check and, where it matters, to verify against the original afterwards. 3D printing has its own tolerance on top of whatever the scan gives you, which is why we treat the scan as the starting point for a part that is then test-fitted, not as a guarantee on its own. If you have a required tolerance, give it to us up front and we will tell you honestly whether scanning is the right way to hit it, or whether hand measurement of a few key dimensions is better.

File formats: mesh or CAD is the real choice

A raw or cleaned scan comes back as a mesh, typically STL, OBJ or PLY. A mesh is a faithful record of the surface, flaws and all, and it is enough if you want to print a copy, check a shape against a reference or keep a record. What you cannot easily do with a mesh is edit it. You cannot change a hole diameter, add a fillet or send it to a machine shop and expect clean toolpaths.

For that you need the scan rebuilt as a parametric solid model, delivered as STEP, which opens in any CAD package. This is a separate stage of work, done through our CAD modelling service, and it is quoted separately from the scan itself. A simple way to decide: if the end goal is a printed copy, a mesh is usually enough. If the end goal is changing the part, making it in another process or keeping an editable master, you want STEP.

A worked example (representative)

This is a representative job, not a specific customer. A maintenance team has a cast aluminium housing from a discontinued pump, with a cracked mounting ear and no drawing. They post us the part. We lightly coat the shiny casting, scan it, clean the data and measure the mounting bores and flange face directly. Because they want to machine a replacement as well as print a trial, we rebuild it as a STEP model, print a test fit in SLA resin to check it against the pump, and adjust before anything is committed to metal. The same measure-print-check discipline is behind our ribber dial case study, where a half-millimetre error was enough to stop a hundred-year-old mechanism meshing.

The point of the example is the order of work. The scan reduces guesswork, the test print exposes what the scan could not, and the final file is only as good as that check.

What drives the cost

Scanning is quoted per job, because the same-sized part can take twenty minutes or a morning. The factors are the ones above: complexity, surface (a part that needs coating or markers takes longer), size and quantity, and, most of all, what you want back. A scan delivered as a mesh is the smaller job. A scan rebuilt as an editable model is a larger one, and it overlaps with reverse engineering, which is where the modelling hours sit. We give a written estimate before any work starts, so you can see what is included and choose the cheaper route if a mesh will do.

Practically, turnaround for a scan, clean-up and printable file is typically five to ten working days. You can bring the object to our Bicester workshop or post it; we handle courier collection within reasonable range. Photos and a rough set of dimensions are enough to get a first view on whether it is worth scanning at all.

If you have a part you need captured, send photos, rough dimensions and what you want to do with the result. We will tell you honestly whether scanning is the right route and give you a written estimate.