Sourcing
Value Engineering: Taking Cost Out of a Part
The short answer
Value engineering removes cost from a part without removing the function it has to deliver. On a machined part most of the available saving sits in how the part is made rather than in what it does: tolerance relief, fewer setups, standard stock sizes and eliminating secondary operations usually return more than material substitution. The ranking matters, because teams reliably start with the material, which is often the smallest lever available to them.
What value engineering is, and what it is not
Value engineering is the disciplined removal of cost that is not buying function. It starts from a working design and asks, feature by feature, what each one is for and what the cheapest way to deliver that function would be.
It is not asking your supplier for a discount. A discount moves margin between two companies and changes nothing about the part. Value engineering changes the part or the process so that it genuinely costs less to make, which is a saving that survives, scales with volume, and does not depend on anyone's goodwill.
It is also not cheapening the product. The discipline is explicitly about holding function constant. If a change degrades what the part does, it is a cost reduction decision with a tradeoff, and it should be argued on those terms rather than smuggled in under a heading that implies it is free.
The distinction matters when you present the results internally. Engineering will resist a proposal that looks like quality being traded away and will engage with one that identifies cost buying nothing.
Where the cost actually sits
You cannot remove cost you cannot see. On a machined part the intuition that material dominates is usually wrong, and it sends teams to the smallest lever first.
The shares below are typical for a low to moderate volume machined part. On high volume parts, setup amortizes away and material becomes proportionally larger; on complex prototype parts, machining time is even more dominant.
| Cost element | Typical share, machined part | What moves it |
|---|---|---|
| Machining time | 40 to 60 percent | Tolerances, features, number of operations, material machinability |
| Setups and fixturing | 10 to 25 percent | How many orientations the part must be held in |
| Raw material | 10 to 25 percent | Alloy, stock size, how much is cut away |
| Secondary operations | 5 to 20 percent | Finishing, heat treatment, plating, masking |
| Inspection and documentation | 2 to 15 percent | Number of critical characteristics, required reports |
| Programming and first article | Amortized | Spread over the run; dominant at quantity one |
The levers, ranked by payback
The ordering below reflects what usually returns most for the least engineering effort and requalification risk. Work down it rather than starting with whatever is easiest to change in CAD.
| Lever | Typical saving | Engineering effort | Requalification risk |
|---|---|---|---|
| Relax non critical tolerances | 10 to 30 percent | Low, a drawing change | Low |
| Reduce number of setups | 10 to 25 percent | Medium, may need feature relocation | Low to medium |
| Remove or simplify unused features | 5 to 20 percent | Low | Low |
| Move to standard stock size | 5 to 15 percent | Low | Low |
| Loosen surface finish requirements | 5 to 15 percent | Low | Low |
| Consolidate or delete secondary operations | 5 to 20 percent | Medium | Medium |
| Substitute material | 5 to 20 percent | Medium, needs analysis | Medium to high |
| Replace custom features with standard hardware | 5 to 15 percent | Medium | Medium |
| Commit to volume or a blanket order | 5 to 15 percent | None | None |
| Change process entirely, for example machining to casting | 30 percent plus | High, plus tooling | High |
Tolerance relief is the biggest single lever
Tolerances are where cost hides, because they are invisible on a rendered model and expensive on a machine.
Cost does not rise smoothly as a tolerance tightens. It steps. At some point the existing process can no longer hold the band, and the shop has to add a finishing pass, move the feature to a grinder, hold the part in a different fixture, or inspect every piece. That step is where a price doubles, and it is invisible unless you ask which tolerance triggered it.
The practical method is to go through the drawing and separate dimensions into three groups: the ones that control a fit or an interface, the ones that control appearance, and the ones that got a tolerance because the template applied one. The third group is usually the largest, and moving it to general tolerance costs nothing and saves real money.
For the ones that do matter, state what they are for. A shop that knows a bore is a bearing seat can propose a fit callout, which is often looser than the number a designer would pick and functionally better. A tolerance with a stated purpose is negotiable. A tolerance without one has to be held exactly as drawn.
Ask your supplier a specific question: which three tolerances on this part are driving the process, and what would each be worth if relaxed. Most shops can answer immediately, because they already know.
Setups multiply, and nobody sees them in CAD
Every time a part has to be unclamped, reoriented and re-clamped, you pay for fixturing, for the labor, and for a new opportunity to introduce error. Setups are also where tolerance stacks between features get worse, because features machined in separate setups cannot hold position to each other as tightly as features machined in one.
A part with features on five faces needs to be held five ways, unless it can be redesigned so several of those features share a face. Moving a mounting hole from a side wall to the top face, giving up a cosmetic chamfer that exists only on the back, or accepting a slightly different bracket orientation can each remove a whole setup.
The saving is not linear. Removing one setup from a five setup part takes out its fixturing, its handling, and its share of the cycle, and it often tightens the achievable relationship between the remaining features as a side effect. This is the lever where an hour with your supplier's engineer returns most, because they can see the setups and you cannot.
Related: deep pockets, thin walls and sharp internal corners each force slower tooling or extra operations. An internal corner radius sized to a standard end mill rather than to an arbitrary value is free money, and it is one of the most common single fixes.
Material substitution without changing performance
Material is where most teams start, and it is usually a mid ranked lever rather than the top one. It is also the change with the highest requalification risk, because it can affect strength, corrosion behavior, thermal properties and appearance simultaneously.
The substitutions that work are the ones where the property you are paying for is not the property the part needs. Moving from 7075 to 6061 where the part is not strength limited. Moving from 316 to 304 where there are no chlorides. Moving from stainless to anodized aluminum where the environment is benign and the mass saving is a bonus. Moving from a hard to machine alloy to a free machining variant where the corrosion performance is not the point.
The substitution that usually disappoints is chasing a cheaper raw material on a part where machining time dominates. Saving 20 percent on a material that is 15 percent of the cost returns 3 percent, and if the cheaper material machines more slowly you can lose more than you saved.
Check stock availability before committing. A theoretically cheaper alloy in a size nobody stocks costs you lead time on every order, and that recurring cost outlives the one time saving.
Secondary operations and the paperwork nobody priced
Secondary operations accumulate quietly. A finish here, a heat treatment there, a plating requirement inherited from a previous product, an inspection report specified once and never revisited.
Go through them and ask what each is for. Finishing on internal surfaces nobody sees. Plating on a part that lives inside a sealed enclosure. Masking on features that could be machined after coating instead. A surface finish callout on every face when only the sealing face needs one.
Documentation deserves the same audit. Every characteristic you designate as critical creates an inspection obligation, and on a part with a capability requirement it creates a capability study. Designating thirty critical characteristics rather than the four that matter can add more cost than any material change will save, and it makes the part slower to quote as well.
This is the category where the answer is most often that a requirement exists because it existed on the last part, which is not a reason.
When a redesign costs more than it saves
Value engineering has a break even, and past a certain point in a program the arithmetic turns against you.
Count the full cost of the change, not just the engineering hours. A drawing revision is cheap. A change that invalidates a completed qualification test is not. A change to a molded part after tooling is cut means modifying or replacing the tool. A change to a part in a regulated product can mean a design control update and a regulatory submission. A change to a PPAP approved part means a resubmission, which has its own cost and lead time.
The rough test: multiply the per part saving by the units you will actually build over the remaining product life, then compare that against engineering time, requalification, tooling changes and the schedule risk of doing it now. On a part at 200 units a year with two years left, a two dollar saving does not fund a requalification. At 50,000 units a year it funds almost anything.
The corollary is that timing dominates. The same change is nearly free before the design freeze, moderately expensive after tooling, and sometimes impossible after qualification. If a program is heading for volume, the value engineering pass belongs before tooling, not after the first invoice arrives and somebody notices the price.
What to send, and what to ask
Send the current drawing, the current price and quantity, and your realistic annual volume. Then send the thing suppliers almost never receive: what the part actually has to do, which dimensions control interfaces, and which requirements come from a customer or a regulator and therefore cannot move.
That last distinction saves everyone time. A supplier who knows which constraints are genuinely fixed will propose changes to the ones that are not, instead of a list you have to reject.
Then ask four questions. Which three tolerances on this part are driving your process, and what is each worth relaxed? How many setups is this part, and which feature would have to move to remove one? Which of the secondary operations here contributes most to the price? And at my volume, is there a different process entirely that changes the economics?
Those four questions, asked of a shop that is already quoting the part, produce a better cost reduction list than most internal reviews do, because the answers come from the people who can see the setups.