DFM — the 30-minute review that pays for itself

Most parts arrive at our shop with at least one feature that costs more than it should. Sometimes a tolerance that's tighter than the part actually needs. Sometimes a bend that's hard to form. Sometimes two materials specified for one part that could be a single piece. None of this is anyone's fault — designers can't know every constraint of every shop. That's why we always ask for a 30-minute review before we cut.
What we actually look at
A DFM review at Deltisan is not a long process. It's mostly four questions:
- Can we hit the tolerances on the equipment we have, or do we need to outsource a step? Outsourced steps cost more and slow down the schedule. If a non-critical dimension is at ±0.05 mm but the function would be fine at ±0.2 mm, that's a free saving.
- Are the bends, holes, and features arranged so the part can be made in one setup? Multi-setup machining is where time disappears.
- Is the material the right choice for the job? Sometimes a cheaper material with the same finish does the same job. Sometimes a slightly more expensive material lets us skip a process.
- Is the finish necessary in the way it's specified? Powder coat where paint would do, or stainless where coated mild steel would last just as long, is common.
A small example
A customer recently sent a sheet-metal bracket with a 0.5 mm bend radius. The drawing didn't say why. We asked. Turns out it was a default in the CAD package, not a functional requirement. Changing it to a 1.5 mm radius matched our standard tooling and cut the per-unit price by about 12% — without changing the function of the part.
That conversation took ten minutes. The customer's order was 800 units. You can do the math.
The bottom line
If you have a drawing that's nearly ready to send out for production, send it to us first. We'll look at it for free. We're not going to redesign your part — we'll just tell you which lines on the drawing are costing you money for no reason. Half the time, you'll change them. The other half, you won't, because there was a reason. Either way, you find out before you commit to a run.