Parts that print first time are designed for the process, not adapted to it. This is what our engineers check before a job goes on a machine — the handful of decisions that decide whether a part comes out right, and the ones that quietly cost you a reprint.
Design for additive manufacturing · DfAM guidelines · escape holes and wall thickness · print orientation

Most design mistakes in additive come from applying one process’s habits to another. The single biggest split is whether your part is built in powder or built with supports, because everything downstream follows from it.
No supports · near-isotropic · nests densely
Supports needed · direction matters · larger and finer
None of these are exotic. They are the things that come up again and again on real jobs, and catching them at CAD stage is free where catching them after a build is not.
Any sealed cavity in a powder-bed part finishes the build packed solid with unsintered powder. If there is no route out, it stays there — adding weight, rattling, and in a fluid part, contaminating whatever runs through it.
Two escape holes is the working minimum: one for powder to leave through, one to let air in behind it. More is better, and bigger is better than smaller. Put them where a cleaning tool can physically reach, not just where they look tidy on the model — a hole into a deep blind pocket is not much use if nothing can get down there to agitate the powder.
If a cavity has to stay sealed for the part to work, say so at quote and we will talk through whether it is printable as designed or better split and bonded.
Sintering and fusing are heat processes, and heat leaves a thick section more slowly than a thin one. Where a wall jumps abruptly from heavy to light, the two sides cool at different rates and pull against each other. That shows up as warp, as internal stress, or as a part that measures fine on the bench and moves later.
The fix is usually to shell a solid part down to a consistent wall rather than printing it solid, and to blend transitions with a fillet or a taper instead of stepping straight from one thickness to another. You get a straighter part, and because there is less material in it, a cheaper and faster one.
Ribs are the honest way to put stiffness back into a shelled part — they add strength without adding a thick section.
With no supports to dig out from between components, a powder-bed build can produce a working mechanism in one go. Hinges, chains, captive nuts, ball joints, enclosed mechanisms — all of it can come off the machine already assembled and moving, with no fasteners and no assembly labour.
The whole thing turns on the clearance between the faces that move against each other. Too tight and the surfaces fuse into a solid lump that no amount of cleaning will free. Too loose and the joint is sloppy. It also has to be a gap powder can actually be cleared from, which ties it back to rule 1.
Tell us what has to move, how freely, and in which grade, and we will confirm the clearance before it goes on the machine rather than after.
An FDM part is built from layers bonded to one another, and that bond between layers is weaker than the material within a layer. Load a part along the layers and it is strong. Load it across them, and you are testing the weld between layers instead of the plastic itself.
This is not a slicer setting to be fixed later — it is a design input. A bracket that will see load in one direction should be oriented so the layers run with that load, and if the geometry will not allow it, that is worth knowing before the part is made rather than after it fails.
Send the load path with the part — even a sketch with an arrow on it — and we will orient to it. On SLS and MJF this matters far less, which is exactly why parts loaded in several directions at once tend to end up on those processes.
The cheapest habit in design for additive. When a part has one critical fit — a snap-fit, a dovetail, a bearing seat, a threaded boss — do not discover it is 0.2 mm out by running the whole part first.
Isolate that feature onto a small test coupon, print the coupon, check the fit in your hand, then commit the full part with the clearance corrected. A coupon is a fraction of the cost and turns round fast, and it removes the one variable most likely to force a reprint.
This matters most where the fit is against something we did not print — a bought-in bearing, a standard fastener, an existing housing. Send both and we will check the mating dimension with you.
On powder processes you pay for the volume of the build your parts occupy, so parts that nest well are parts that price well. A batch that packs tightly into a bed costs meaningfully less per part than the same batch ordered piecemeal across separate builds — which is worth knowing before you split an order.
If the part is larger than the envelope, that is not the end of it. We split it, print it in sections, then bond and finish in-house. Send the overall dimensions with the enquiry and we will plan the split so joints land somewhere sensible rather than through a critical feature or a show face.
And if you are ordering repeats over time, tell us — consistency across builds is easier to hold when we know from the start that it matters.
The numbers that will not bend. Anything larger than these is printed in sections, bonded and finished in-house.
| Process | Max build size | Lead time |
|---|---|---|
| SLS | 165 × 165 × 300 mm, up to 490 × 490 × 740 mm for PA12-CF | 1–3 days express, 5–7 days standard |
| FDM | 500 × 500 × 900 mm (340 × 320 × 340 mm for PPA-CF) | 1–3 days express, 5–7 days standard |
| SLA | 353 × 196 × 350 mm | 1–3 days express, 5–7 days standard |
| MJF | 380 × 284 × 380 mm | 5–7 working days, standard only — no express tier |
| Tolerance | ±0.3 mm up to 100 mm, then ±0.3 mm for the first 100 mm plus ±0.1% for every 100 mm thereafter. Full detail on tolerances & accuracy. | |
Tighter fits are achieved by printing near-net and machining the critical features — flag those dimensions at quote. Send the CAD and we will confirm the envelope against your part.
FDM behaves differently enough from the powder processes that it is worth calling these out separately.
Before the print
After the print
Minimum wall thickness, hole diameter, embossed text height, pin diameter and clearance all move with the process and with the specific grade. One set of numbers across twenty-four production materials would be tidy, and it would also be wrong for most of them — a figure that holds for SLS PA12 does not hold for FDM PPA-CF or SLA Ceramic-Filled.
So we publish them by process. The SLS design guide gives the full set of reference dimensions for powder-bed nylon — walls, pins, holes, drain holes, clearances, lattice gaps and embossed and engraved text minimums — plus grade-specific notes for PA12, PA11, PA12-GF, PA12-CF and TPU 90A. Guides for the other three processes are in preparation.
For anything not yet covered, send the part and tell us the grade, or let us pick the grade, and we will confirm the real minimums for that combination before you design around a guess. It is free, it comes with the quote, and it takes an engineer rather than a table.
Every quote comes with a free design-for-additive review — we tell you what will print, what will not, and what we would change. Printed, finished and inspected under one UK roof.
Printed, finished and inspected at our own ISO 9001 facility in Shipston-on-Stour, Warwickshire. Call us on 024 7736 0144 or request a free sample pack.