Material testing means measuring a property on a sample so that a part can be trusted without being broken, and the useful habit is to make a test coupon from every melt, pour or heat treatment run and test it with the cheapest method that answers the question. Pulling a bar gives strength and stretch, an indentation gives hardness, penetrant and magnetic powder find surface cracks without damage, and a polished and etched section shows the structure behind the numbers. A test result only means something if it is recorded with the batch it came from.
Part of Materials Science
Testing and quality control is the practice of verifying material properties by destructive and non-destructive tests, and of linking each result to the batch and process that produced it.
Test The Question, Not The Part
Each test answers one question, and the right choice depends on what you need to know and whether the part must survive.
| Question | Method | Destroys the sample? | Article |
|---|---|---|---|
| How strong, and how much does it stretch? | Pull a bar to failure | Yes | Tensile Testing |
| How hard is the surface? | Indent or file | Slightly | Hardness Testing |
| Are there cracks? | Penetrant or magnetic powder | No | Non-Destructive Testing |
| What is the structure? | Cut, polish, etch, look | Yes (a slice) | Metallography |
The Advanced Materials article lists the same set. Note that the quality-testing articles in other parts of the corpus are about fired bricks, rubber and penicillin respectively, not about metals, so this section does not use them.
Make A Test Coupon With Every Batch
The part that goes into service cannot be broken to prove it is good. So cast, forge or heat treat a small coupon alongside it from the same material in the same run. If the coupon passes, the batch is trusted. A coupon is the only honest link between a test result and a part:
- Mark it with the batch number the moment it is made.
- Treat it the same way, in the same furnace, at the same time and in the same quench.
- Make it representative, with a thickness similar to the part, because thin sections cool faster than thick ones.
- Keep the record. Write down the batch, process, date, and result.
Know What Each Number Means
Strength figures, from Mechanical Properties, are not safe working loads. The working load is the failing load divided by a safety factor, as Safety Factors defines, and the factor should be larger when your testing is crude. A homemade test rig that reads within 10% justifies a bigger margin than a calibrated one.
Quality Is A Habit Of Repeating
One good result does not make a process reliable. Repeat the test on several coupons. If results scatter, the process is not controlled, whatever the average says. Fix the scatter first by tightening temperatures, times and weights.
Testing stores energy
A loaded test rig, a pressurised cylinder and a part about to break all hold energy that releases suddenly. Put a shield between you and the specimen, stand to one side and never lean over a rig under load.
Common Mistakes
- Testing the wrong thing — hardness says little about toughness.
- Testing once — a single pass shows luck as easily as quality.
- Losing the label — a result without its batch is useless.
What’s Next
- Learn next: Tensile Testing — the core strength test
- Combine with: Non-Destructive Testing — checking the real part