What a DFM review is, and when it should happen
A design-for-manufacture review is the pass a drawing goes through before anyone cuts metal, where the shop compares what the drawing asks for against what the process can actually hold.
The timing is the whole point. A change made on a drawing costs nothing. The same change made after the die is cut costs the die. The same change made after the parts ship costs the relationship.
We run this review before quoting rather than after winning the order, which is why our quote takes 24 hours rather than minutes. This page lists the things that review most often catches, so you can find them in your own drawing first.
1. A bend radius that is below material thickness
The single most common flag. When the inside bend radius is smaller than the material thickness, the outer surface is being asked to stretch beyond what the material tolerates, and the part shows it as cracking, orange peel or a fracture line at the bend.
Stainless and high-strength grades are less forgiving than mild steel. The fix is usually to open the radius, or to accept a relief, or to change the material. Note that this is a forming limit, not a design preference — no amount of press tonnage changes it.
2. A hole too close to a bend line
When a hole sits within roughly 1.5 to 2 times the material thickness of a bend line, the hole distorts during forming. It may still be the right diameter on the drawing and wrong in the part, which is the harder failure to catch, because the inspection report and the drawing will agree with each other while the part still will not assemble.
The fix is to move the hole, add a relief slot, or form first and punch after.
3. A tolerance the process cannot hold, applied to a dimension that does not need it
A drawing that applies a tight tolerance to a dimension with no functional consequence is more expensive than it needs to be, and nothing is gained.
Different processes hold different tolerances. On our routes, laser cutting holds about ±0.05 mm, forming about ±0.1 mm, and five-axis machining ±0.005 mm on the features that require it, verified on a CMM with a stated measuring accuracy of ±0.002 mm. A part that carries a machining tolerance across a forming operation cannot be made as drawn, and the review catches that before it becomes a rejection.
4. A flat pattern that will not close
Some geometries cannot be formed from flat sheet without an additional operation. A deep draw with a small corner radius, a return flange with no access for tooling, a closed shape with an undercut — each of these needs either a change to the design or an honest warning about what the forming sequence will look like.
The review should tell you which, and what the alternative costs.
5. A weld joint that cannot be reached, or a fit-up that has no datum
Welding is where tolerances compound across a fabrication. Two parts each held to their own drawing can still produce an assembly that will not close, because nobody owned the accumulated tolerance.
The review asks where the datum is for the assembly, whether the fixture can reach the joint, and whether the weld sequence will pull the part out of square. A weld jig may be the cheapest part of the whole project, and it is the thing most often left out of a low quote.
6. Surface finish applied before the operation that would ruin it
Powder coating, plating and anodising are usually last, because welding burns coating off and machining cuts through it. A drawing that implies a finished surface before assembly has the sequence wrong, and the correction changes the cost.
Where a finish is genuinely required between operations, that is worth stating explicitly, along with what the subsequent steps are allowed to do to it.
7. Material choice that is expensive without being necessary
Grade 316 stainless where 304 performs the same function, or a tight thickness specification where standard tolerance would do, adds cost without adding capability.
Conversely, a material that is cheap and wrong is worse. The review should question both directions, and it should say which properties are actually load-bearing in the decision — corrosion, weldability, formability, finish.
8. A quantity that makes the process choice wrong
The cheapest route depends on the number. Flat profiles are laser cut; parts at higher volume move onto stamping once the tooling cost is justified by the quantity. A design optimised for one route at 50 pieces may be the wrong design at 50,000.
The most useful thing a review can do here is tell you where the crossover is, and what changes at that point.
What the review output looks like, and what it does not claim
| The review gives you | The review does not claim |
| A process route with a tolerance at each stage | That the part will work in your assembly |
| Specific flags with the reason attached | That every risk has been found |
| The cost effect of each suggested change | That a change is necessary rather than advisable |
| The volume where the process route changes | A guarantee on final dimensions before first article |
| An honest list of what cannot be made | Approval in regulated sectors we do not hold certification for |
Two things to keep in mind. First, a review is engineering advice based on the drawing you supplied; if the drawing is incomplete, the advice is incomplete. Second, the review is not a substitute for first article inspection, which is where the part is measured against the drawing for the first time.
How to get the most out of a review
- Send the 3D model and the drawing together. The model shows geometry, the drawing says which dimensions matter, and neither is complete alone.
- Mark the functional dimensions. A review that knows what moves and what merely locates is a much better review.
- State the quantity, including the annual figure if you know it. The process choice depends on it.
- Ask what they would change. A supplier who cannot suggest a single change has not looked.
If you are trying to work out whether a supplier has an engineering function at all, the signals are set out in telling a real sheet metal factory from a trading company. The full vetting list is in the twelve-point supplier checklist.
Key facts
| Quote turnaround | 24 hours, including a design-for-manufacture review |
| Typical tolerances | Laser cutting ±0.05 mm, forming ±0.1 mm, five-axis machining ±0.005 mm on the features that require it |
| Measuring method | Coordinates measured on a CMM with stated accuracy of ±0.002 mm |
| Bend limit | Inside radius below material thickness is a forming limit, not a preference |
| Hole-to-bend clearance | Below roughly 1.5 to 2 times material thickness the hole distorts |
| Sample lead time | 7 days typical |
| Minimum order quantity | None |
Frequently asked questions
What is a DFM review in sheet metal fabrication?
It is the check a drawing passes through before production, comparing what the drawing asks for against what the process can hold. It runs before quoting rather than after, so that changes are made on paper where they cost nothing instead of on a die where they do not.
When should I ask for a DFM review?
Before you commit to tooling, and ideally before you finalise the design. The earlier it happens the less it costs, because a change on a drawing is free while the same change after a die is cut is a tooling project. Sending the model and drawing together gives the most useful result.
What is the most common problem a DFM review finds?
A bend radius below material thickness. The outer surface is then asked to stretch further than the material tolerates, which shows up as cracking or a fracture line. Stainless and high strength grades are less forgiving than mild steel, and no amount of press tonnage changes the limit.
Does a DFM review guarantee the part will work?
No. It is engineering advice based on the drawing supplied, so an incomplete drawing produces incomplete advice, and it does not confirm that every risk has been found. It is also not a substitute for first article inspection, where the finished part is measured against the drawing for the first time.
Related reading
