What perishable tooling costs per piece
Perishable tooling is the line most floors carry as a percentage, usually somewhere between 2 and 5 percent of machining cost, applied without anyone having checked it in years. On a high volume turned part it is a real number and it can be calculated properly in about 10 minutes.
This post does that arithmetic on one operation. All figures are illustrative and the point is the method rather than the values, because tool life on your machines in your material is the one input nobody else can give you. The figure it produces belongs on every operation of a machined part build-up.
What perishable tooling actually covers
Anything consumed by cutting rather than owned by the job. Inserts, solid carbide drills and end mills, taps, reamers, grinding wheels, saw blades and the dressing sticks that maintain them.
The boundary that matters for quoting is that this is per piece cost, not per program cost. A dedicated form tool built for one part is tooling that gets amortised across a volume. A standard insert that wears out and gets replaced is a consumable that attaches to every piece you cut.
Holders, collets, chucks and boring bars sit on the other side again. They belong to the work center rather than to the job, which is why they appear in the machine’s annual cost rather than in a per piece line.
Getting those 3 categories separated is most of the work. Once they are, each one has a natural home and nothing gets counted twice or missed entirely.
The arithmetic on one operation
Take a turning operation on the worked part and cost the insert properly. An illustrative CNMG insert costs $9.40 and has 4 usable edges, so each edge costs $2.35.
At an illustrative tool life of 400 pieces per edge, the insert cost per piece is $2.35 divided by 400, which is $0.0059. Under six tenths of a cent, which sounds like nothing and is the reason nobody checks it.
Now add the tool change. Indexing to a fresh edge takes an illustrative 3 minutes of machine time including the check piece. At the $89.61 hourly rate from a rate built from the ground up, 3 minutes is $4.48.
Spread across the same 400 pieces that is $0.0112 a piece, which is nearly twice the insert itself. The total perishable cost for this operation is $0.0171 a piece, and two thirds of it is machine time rather than carbide.
The change costs more than the edge.
Why the cheapest insert is often the most expensive line
Halving the insert price while halving tool life makes the line worse, not better. Run the same operation on an illustrative $5.20 insert with 4 edges that lasts 200 pieces per edge.
Edge cost is $1.30, so insert cost per piece is $0.0065, slightly worse than the expensive one despite costing 45 percent less to buy. The tool change now happens twice as often, so the change cost per piece doubles to $0.0224.
Total is $0.0289 against $0.0171, which is 69 percent more expensive on an insert that looked 45 percent cheaper in the catalogue.
Push it the other way and the effect reverses. An illustrative $16.80 insert lasting 800 pieces per edge gives $0.0053 of carbide and $0.0056 of change time, totalling $0.0109, which is the cheapest of the 3 while being the most expensive to buy.
That is the whole argument for letting the people running the machines choose tooling. The purchasing view optimises the wrong number, and the difference across these 3 options at 1,200,000 pieces a year is $21,600.
What a tool change really costs on your floor
Machine minutes are the expensive part, so the tool change deserves its own estimate. Three minutes is a reasonable figure for indexing an insert on a running job. It is not a reasonable figure for changing a boring bar that needs re-setting, which can be 20.
Unattended running changes the sum again. A machine running lights out that stops for a tool change at 2am does not resume until somebody arrives, so the cost of that change is not 3 minutes of cycle time, it is 6 hours of an idle machine.
That is why tool life is a scheduling input as well as a cost input. Matching tool life to a shift length, or to the bar magazine capacity, is worth more than the carbide saving on almost any high volume job.
Tool life is a scheduling number.
Predictability matters as much as duration. A tool that reliably lasts 380 pieces is more useful than one averaging 450 with a wide spread, because the first can be changed on a planned break and the second fails somewhere in the middle of a run and takes a piece with it.
Sister tooling changes it in the other direction. A machine that can index to a duplicate tool automatically turns a 3 minute stop into a few seconds, at the cost of a second tool sitting in the turret and a position you may not have spare.
Regrinds, and when they are worth it
A regrindable tool has a different cost structure and needs a different sum. A solid carbide drill at an illustrative $84 might be reground 4 times at $18 a regrind, giving 5 lives for $156 rather than 5 tools for $420.
The catch is that a reground tool is often shorter, sometimes smaller in diameter, and rarely as good as new on the first pieces. If your process tolerance cannot absorb that, the regrind is not a saving, it is a scrap risk on a part that has already absorbed 6 operations.
Logistics carry a real cost too. Tools leave the building, come back in 2 weeks, and somebody manages the float. That is inventory and administration against a saving that on a low volume part may be a few hundred dollars a year.
The honest rule is that regrinding pays on high volume, repetitive work with tolerance headroom, and does not pay on short runs. Deciding it once per tool family and writing it down beats deciding it per purchase order.
A reground tool is not a new tool.
Grinding wheels sit in the same category and are usually forgotten entirely. A wheel has a life, a dressing frequency and a dressing cost, and the diamond that dresses it is itself a consumable. On a part with a ground feature that chain is a real per piece line rather than an overhead.
Getting it into the quote without a percentage
Carry perishable tooling as a per piece figure on each operation, derived the way the sum above derives it. It takes an hour to build for a work center and it stays valid until the material or the tooling changes.
Where a percentage is genuinely the practical answer, on low volume work with many small operations, derive the percentage from a few worked examples rather than inheriting it. A figure that came from somewhere is defensible. A figure everybody has always used is not.
The burdened rate should not already contain perishable tooling if you are also carrying it per piece, and this is the most common double count in estimating. Check which side of the line your rate was built on before adding a second one.
Then compare it against what the job actually consumed. Tool life is the input most likely to be optimistic, and the only place the truth exists is the crib issue records against the pieces produced.