How to Identify an Unknown Metal
Start with markings and part history, then combine magnetic response, measured density and surface appearance to narrow the candidates. These checks can suggest a metal family, but they cannot reliably establish an exact alloy or heat-treatment condition. When the decision depends on grade, use suitable positive material identification (PMI) or laboratory testing that measures the defining elements and properties.
Decide what you need to identify
Separating aluminum-like scrap from steel-like scrap is a different task from confirming 6061-T6 plate or a low-carbon stainless grade. Before testing, write down the decision: preliminary sorting, machining, welding, supplier acceptance or return to service. The more the decision depends on a specific composition or condition, the less useful a visual guess becomes.
Set an unverified production part aside, assign it an ID and photograph its markings. Check the drawing, purchase record, heat number and material report for a reliable link to that piece or lot. Also inspect its construction: a coating, steel insert, hollow section or cladding can make the outside behave differently from the bulk material.
ASTM E1476’s published scope and guidance distinguish methods that measure composition from methods that respond indirectly to material properties. Select the method to resolve the actual identification problem.
Inspect before creating dust or sparks. Do not grind, drill, heat or chemically test an unidentified coating, sealed item or possible reactive metal without an appropriate assessment and procedure. Preserve critical or valuable parts until a permissible sampling method is agreed. Metal powder can behave very differently from a solid part; the CSB’s titanium-powder investigation documents that distinction’s practical importance.
If an item carries a radiation warning or gives reason to suspect a radioactive source, stop handling it and contact the site safety lead and relevant radiation authority. EPA guidance for suspected radioactive scrap explains the response.
Read magnetic response and color as clues
Use the same magnet and compare several locations: flat areas, bends, edges and any inserts. Record where the pull changes. Then inspect whether the visible color belongs to bare metal, a coating or corrosion.
A magnet does not identify “mild steel”
Strong attraction is common in carbon steel, cast iron and ferritic, martensitic or duplex stainless steel. Nickel and cobalt can also attract a magnet. A magnetic result therefore identifies a magnetic constituent, not one commercial grade. See BSSA’s discussion of magnetic sorting.
No obvious pull does not rule out steel. Annealed austenitic stainless can show little attraction; cold working can form magnetic martensite, creating pull around bends or machined areas. BSSA explains this change in magnetic response. Coatings, part thickness and a hidden insert can also affect the observation.
Separate surface color from substrate color
Red-orange bare metal supports a copper-rich candidate; yellow may suggest brass or another copper alloy. Neither color establishes a grade. Silver-gray leaves many candidates open, while paint, plating and oxide can conceal the substrate completely.
For a yellow fastener with strong magnetic attraction, investigate a plated magnetic core or an insert before calling it solid brass. Use existing exposed features first; create a fresh witness area only if surface alteration is permitted and its hazards are understood.
Measure density with a volume you can trust
Density is mass divided by material volume: ρ = m / V. A scale gives mass; suitable dimensions or displacement give volume. With grams and cubic centimeters, the result is g/cm³. One milliliter equals one cubic centimeter.
For a simple solid block, use its measured length × width × thickness. Do not treat the outside of a tube as a solid cylinder. Hollow parts, pores, attached hardware and multiple materials can turn the result into an apparent average rather than the density of one metal.
A small reading error can change the comparison
Suppose a 54 g sample raises the level from 40 to 60 mL. The nominal result is 2.70 g/cm³. If each water reading could be wrong by up to 1 mL, the displaced volume could be 18–22 cm³. Ignoring mass error for this illustration, the density could then be about 2.45–3.00 g/cm³. Extra decimal places in a calculator cannot recover missing measurement precision.
Repeat the readings and check whether the uncertainty is small enough to separate your candidates. A density near an aluminum reference can support a family-level screen; it cannot prove 5052 versus 6061, a casting grade or temper.
| Element | Reference density, g/cm³ |
|---|---|
| Magnesium | 1.74 |
| Aluminum | 2.70 |
| Titanium | 4.54 |
| Zinc | 7.13 |
| Iron | 7.87 |
| Nickel | 8.90 |
| Copper | 8.96 |
| Lead | 11.35 |
Rounded from NIST’s elemental-media constants. These are reference values for elements, not allowable ranges for alloy families or a complete list of candidates. Composition, temperature and sample construction affect comparisons. In particular, the iron entry is not a specification for every steel.
Use visual spark testing only as a controlled comparison
A trained operator may compare spark streams from suitable ferrous samples against known references to help sort broad groups. The wheel, contact conditions and surface affect what appears. Spark length, color and branching are not a direct reading of carbon percentage or proof of a steel designation.
The classic 1933 Bureau of Standards study by R. W. Buzzard found value in grouping steels of similar composition, while explicitly limiting the identification of unknown steel. Its sorting observations were tied to the grinding setup used. This is evidence about the method’s limits, not a modern grinding safety procedure.
Only consider a spark comparison after the sample and coating are judged suitable, surface damage is permitted and trained personnel have an approved procedure with the necessary machine, exposure and fire controls. Do not use grinding to discover whether an unknown light metal is safe to grind. If the result is unexpected, stop and choose a more suitable verification method.
Visual sparks and spark OES are different tests. Optical emission spectrometry measures an emission spectrum with an instrument and calibration. Watching a grinder’s sparks does not provide equivalent chemistry data.
Combine clues without turning them into a grade label
Use the observations to eliminate explanations and choose a useful next test. The examples below assume suitable samples and valid measurements. They are possible interpretations, not an exhaustive identification chart.
| Observed combination | What it can support | What to check next |
|---|---|---|
| No obvious magnetic pull; silver-gray solid; density near 2.7 g/cm³ | An aluminum-family candidate. The observations do not distinguish specific aluminum grades or tempers. | Check traceability and use composition analysis appropriate to the candidate alloys; verify condition separately if needed. |
| Little magnetic pull on a flat area, stronger pull at formed corners | Cold-worked austenitic stainless is one possible explanation; construction and geometry still matter. | Compare locations and inspect for inserts or layers. Use chemistry if a stainless grade must be confirmed. |
| Yellow surface with strong magnetic attraction | A magnetic substrate beneath a finish, or mixed construction, deserves investigation. | Identify which layer is being examined. A surface-sensitive result must not be reported automatically as bulk composition. |
| Red-orange bare metal; density near the copper reference | A copper-rich candidate. Appearance and density do not establish purity or a named copper alloy. | Use suitable elemental analysis when composition or value matters. |
| Strong magnetic pull but unexpectedly low apparent density | A hollow section, insert, assembly or volume error may explain the conflict. | Inspect construction and repeat the volume measurement before ranking material candidates. |
On a narrow screen, scroll the table horizontally.
Record a bounded result such as “observations are consistent with an aluminum-family solid; alloy and temper unverified.” If clues conflict, preserve the conflict in the record. Counting several weak observations does not create a meaningful confidence percentage.
Choose PMI for the element or property that matters
Positive material identification uses an appropriate analytical method to check material identity. A software grade match is useful only when the sample, measured elements, calibration and acceptance limits fit the question. Ask for measured results and method details, not just the name displayed on the screen.
For coated, clad or small parts, agree on a representative test area before measurement. The coating may need separate analysis or permitted removal. Use reference materials and quality checks appropriate to the instrument and alloy family. Handheld equipment also needs its manufacturer’s radiation or laser safety controls.
Choose additional physical tests only when they resolve a remaining question. Calibrated conductivity or hardness comparisons can help sort selected materials, but they depend on factors beyond chemistry. Heat-treatment condition, microstructure or mechanical performance may require separate records or tests.
For stainless-specific distinctions, continue with the 304 vs 316 stainless steel identification guide.
Handheld XRF
X-ray fluorescence can measure many alloying elements and distinguish numerous alloy families. Typical handheld XRF does not measure carbon, so it cannot confirm an L or H carbon suffix. Coatings and measurement geometry need attention.
LIBS
Laser-induced breakdown spectroscopy samples a small ablated area. A carbon-capable instrument with a validated method can address carbon-sensitive distinctions. Confirm the actual model, elements, alloy matrix and surface-preparation requirements.
Spark OES
Spark optical emission spectrometry can measure carbon and other elements with a suitable method. It generally requires prepared material, appropriate argon and calibration, and leaves a local test mark.
Targeted laboratory testing
Request the analysis needed for the missing evidence—for example carbon measurement or a condition-related test. Confirm the laboratory’s method scope, representative sampling and reporting limits. A chemistry match alone does not establish temper or mechanical properties.
Instrument principles and carbon capability: Thermo Fisher Scientific’s XRF, LIBS and OES comparison. Capabilities must be confirmed for the selected instrument and method.
Keep the result tied to the part and its intended use
A test is useful only if its result can be traced back to the sampled material. Record the piece or lot ID, markings, photos, construction, tested locations and raw measurements. For instrument testing, retain the method, relevant measured elements, reference checks, limitations and the specification used for comparison.
Example: a “316” reading before welding
Suppose mixed stainless offcuts have no reliable material records. An XRF reading supports a 316-family composition, but the fabrication requirement calls for a verified low-carbon grade. The reading does not resolve carbon. Request an appropriate carbon measurement and review the full material requirements before assigning the required grade. This is a hypothetical decision example, not an Oceanplayer test result.
The XRF carbon limitation is explained in Thermo Fisher’s guidance on using XRF and LIBS.
Do not use one convenient piece to establish the identity of an uncontrolled mixed lot. Agree on the sampling and disposition rules, then label only what the evidence supports. For laser cleaning, welding or marking, carry the verified substrate information and coating details into a representative process trial; identifying the metal does not qualify the laser settings.