Holding Tight Tolerance on Aerospace Aluminium Bracket: A Machining Case Study
How we held the required tolerance on a structural bracket machined from 7075-T6 plate: where the dimensional risk sat, the controls we put in place, how the result was verified and what the customer received at goods-in.
The tolerance requirement
The drawing called for a finished part described as 2 mm walls with a profile tolerance of plus or minus 0.05 mm, which is tight enough that process drift alone can push parts out of tolerance.
The customer had experienced variation with a previous supplier and wanted evidence, not assurances, that the dimension would hold across the batch.
Where the risk sat
Risk came from three places: fixture repeatability, thermal growth during a long cut, and tool wear across the batch.
Each was measured before any attempt was made to compensate in the control.
The controls we used
We used a dedicated fixture with hardened locating pins, an on-machine probe to set the work offset and to check the critical feature, and a defined tool change interval based on part count.
Coolant temperature and concentration were controlled, because a few degrees of drift is enough to move a dimension of this size.
Verification
Every part in the sample plan was measured on a coordinate measuring machine, and the results were reported against the drawing with the measurement uncertainty stated.
The report showed capability on the critical characteristic rather than a simple pass or fail.
Outcome at goods-in
The batch passed incoming inspection with no non-conformances, and the customer removed its own incoming check for the following orders.
The agreed control plan is now the reference for repeat production of a structural bracket machined from 7075-T6 plate.
Summary
This case shows the pattern behind most successful CNC programmes: decide the datums early, prove the process once, then measure what you ship.
