Environmental Testing: Six Elements That Solve 80% of Quality Problems
Every machined lot carries variation. This page breaks that variation into six elements: people, machine, material, method, measurement and environmental testing. Read it to decide which element to investigate first when a dimension drifts, and when to leave the others alone.
In this article
- 1
- 2
- 3
- 4
- 5
- 6
Why the Same Setup Still Produces Different Parts
Two engine brackets cut on the same machine, from the same bar, by the same operator will not measure exactly alike. The gap might be 3 μm on a bore and 15 μm on a shoulder height. That spread is normal. What matters is whether it stays inside the tolerance band on the drawing.
Before you touch a machine, separate the two kinds of variation. Common-cause variation is the noise a process always has. Special-cause variation is a sudden shift you can trace to one thing: a worn insert, a new bar lot, a chiller fault.
A simple gage R&R plus a run chart on three critical dimensions, run across twenty consecutive parts, tells you which kind you have. If the chart drifts, stop and find the cause. If it scatters evenly, the process is stable and the drawing is the problem.
This is where environmental testing earns its place in the 5M1E set. People, machine, material, method and measurement get most of the attention. The sixth element quietly moves dimensions while everything else looks perfect.
Human-Machine-Material Environmental Testing: The Six Elements
The six elements come from 5M1E: man, machine, material, method, measurement and environment. On a CNC floor, each one leaves a different fingerprint. Learn the fingerprints and troubleshooting stops being guesswork.
Man covers operator skill, shift handover and how closely setup sheets are followed. Machine covers spindle condition, ball screw wear, thermal growth and fixture rigidity. Material covers hardness, residual stress and lot-to-lot chemistry.
Method covers feeds, speeds, workholding sequence and tool paths. Measurement covers gage choice, calibration and how the part sits on the granite. Environmental testing covers ambient temperature, humidity, vibration and coolant temperature.
None of these six acts alone. A 4 °C room swing is harmless on a ±0.1 mm bracket and fatal on a ±0.005 mm bore. That is why we rank the six by how much they can move the dimension before we start cutting.
How Environmental Testing Moves a Dimension
Aluminium expands roughly 23 μm per metre per °C. A 4,000 mm beam swinging 5 °C moves about 0.46 mm in length. Cut that beam at 18 °C and inspect it at 26 °C and the part looks out of tolerance when it is not.
Steel moves about half as much, near 11–12 μm per metre per °C. Titanium sits around 8.6 μm. The smaller the coefficient, the more the shop can tolerate a warm afternoon without chasing a ghost.
The fix is not always air conditioning the whole plant. Bring the part and the gage to 20 °C ± 2 °C, let both soak, then measure. A 30-minute soak on a 100 mm aluminium part removes most of the error that a one-minute check would report.
Humidity matters less on metal and more on fixtures, gages and composites. Above 60% RH, cast iron surface plates and some fixture plates start to move. Keep metrology rooms near 45% RH and log it.
When Vibration and Coolant Temperature Matter More
Vibration shows up as chatter, poor Ra and short tool life rather than a clean size error. A surface finish drifting from Ra 0.8–1.6 μm to Ra 3.2 μm on a finishing pass usually means the floor or the fixture is moving.
Check the obvious first. Is the machine on leveling pads? Is a stamping press or grinder running next door? Does the problem appear only on the afternoon shift? Vibration logs over one week beat any single reading.
Coolant temperature is the quiet one. A 10 °C rise in coolant shifts the spindle and the part together, and the first parts after a cold start run oversize. Warm up the spindle for 20–30 minutes before cutting tight bores.
Both effects shrink when the process is set up to finish-cut in one pass with a sharp tool, rather than taking spring passes that let the material relax between cuts.
Ranking the Six Elements Before You Cut Metal
Ranking saves time. For a ±0.1 mm sheet metal bracket, environment rarely matters and method matters most. For a ±0.005 mm bearing housing, environment and machine thermal state sit near the top.
A practical order for tight work: material condition, machine thermal state, method and workholding, environmental testing, measurement, then operator technique. Operator technique matters, but it is rarely the first cause on a stable process.
We confirm material condition with a raw material check before cutting. Hardness and chemistry cards arrive with the bar. If a lot is 15 HB harder than the last one, feeds and speeds change before the first part is scrapped.
For work that has to hold ±0.005 mm across a 4,000 mm part, we cut, soak and inspect in the same temperature-controlled cell. The report ships with the part. Ask for it on the quote.
Which Element to Investigate First
Match the symptom to the element before adjusting anything.
| Symptom | Most likely element | First check |
|---|---|---|
| All parts shift to one side | Material or method | Bar lot hardness and chemistry card |
| Size drifts through the shift | Machine thermal state | Spindle warm-up and coolant temp |
| Scatter widens, no drift | Environment or measurement | Room temp log and gage R&R |
| Ra worsens on finish pass | Machine or environment | Leveling pads and nearby vibration |
| First part out of tolerance only | Method or machine | Workholding and tool offsets |
| Bore oval, not round | Machine | Spindle runout and fixture clamp order |
Fix the One That Moves the Dimension
If the part is small and the tolerance is loose, spend your time on method and workholding. If the tolerance is ±0.005 mm or the part runs past 1,000 mm, control material condition and thermal environment first, because no amount of operator skill will hold a dimension the room is moving.
Questions Engineers Ask
Do we need a temperature-controlled room for every tight part?
No. For tolerances of ±0.05 mm or looser on steel and aluminium parts under about 300 mm, a stable shop at 20 °C ± 5 °C is usually enough.
A controlled cell earns its cost when the tolerance is ±0.005 mm, when the part runs past 1,000 mm, or when the drawing names a measurement temperature. Soak the part and the gage at that temperature before reading numbers.
How long should a part soak before final inspection?
For a 100 mm aluminium part, 30 minutes at 20 °C ± 2 °C removes most thermal error. For a 1,000 mm part, allow 1–2 hours. Mass matters more than length: thick sections hold heat longer.
A simple rule we use is one hour per 100 mm of the thickest section, or until the surface and core read the same with a contact thermometer.
Does environmental testing replace a gage R&R?
No. Gage R&R tells you whether your measurement system can see the variation you care about. Environmental testing tells you whether the shop is adding variation the gage then reports.
Run gage R&R at least once per gage on the parts it measures, and log room temperature alongside the results so the two can be compared later.
Which materials are most sensitive to shop temperature?
Aluminium alloys such as 6061 and 7075 move the most, near 23 μm per metre per °C. Magnesium AZ31B and AZ91D are similar. Titanium TC4 moves about 8.6 μm, and steel around 11–12 μm.
Composites and plastics are a separate case: they absorb moisture and grow, so humidity control matters as much as temperature for those parts.
Can vibration be measured without buying instrumentation?
For a first look, yes. A dial indicator on a magnetic base, read over a few minutes on the machine table, shows the low-frequency movement that comes from nearby presses or grinders.
For chatter and higher frequencies, a handheld vibration meter is enough to compare shifts and days. Log the readings next to surface finish results so the pattern is visible.
What environmental data should ship with a precision part?
We can include the measurement temperature, the gage used and the room condition at final inspection. Raw material cards and in-process checks travel with the same report on request.
If your drawing names a specific measurement temperature, say so at quote stage so the inspection plan is built around it rather than added afterwards.
Send a Drawing and We Will Tell You Which Element Matters
Upload a part and our engineers return a quotation with free DFM analysis within 12 hours, including a note on the elements that will drive your tolerance.
12-hour quote100% inspectionNo minimum order
More CNC Process Notes
We publish setup notes, tooling trials and inspection data from the factory floor.
