DO-160 and MIL-STD-810: what qualification actually involves

Two standards, two philosophies

RTCA DO-160, published in Europe as EUROCAE ED-14 with the same text, is the civil reference: environmental conditions and test procedures for airborne equipment. Edition G is what programmes ask for in practice. Its logic is a catalogue of categories — you declare which category your equipment meets in each section, and the resulting table is what an installer reads to decide whether the box belongs in that bay of that aircraft.

MIL-STD-810, currently at revision H, is the other major reference: environmental engineering considerations and laboratory tests. Its logic is the opposite. Rather than standard levels, it expects you to derive the test conditions from the platform life-cycle environmental profile — where the equipment lives, what it is bolted to, what the mission looks like — and to justify that tailoring. Its methods describe how to test, not what level to apply.

That difference has a practical consequence. DO-160 gives you comparability: two units declaring the same category have been through the same thing. MIL-STD-810 gives you realism: the test matches the platform, and nothing else. Programmes routinely ask for both, which is entirely reasonable — they answer different questions.

Categories and tailoring: the part everyone gets wrong

A DO-160 category is not a quality grade. It describes an installation. Temperature and altitude categories follow whether the equipment sits in a pressurised, temperature-controlled bay or in an unpressurised zone exposed to ambient conditions; vibration categories follow the aircraft type and the zone; and Category X means the section does not apply and no test is performed.

Category X is legitimate and widely misunderstood. An avionics box permanently installed in a pressurised bay does not need a waterproofness category, and declaring one would be misleading rather than impressive. What matters is that the declared categories cover the intended installation — which is exactly why the deliverable is a form, not an adjective.

On the MIL-STD-810 side, the equivalent discipline is tailoring: a vibration test is a chosen spectrum, a chosen duration and a chosen number of axes, and every one of those choices has to be traceable to the platform. "Tested to MIL-STD-810" without the methods, procedures and levels means nothing at all — the standard says so itself.

What the sections cover

DO-160G has twenty-six sections, of which about twenty are tests. They fall into four practical families, and a typical airborne recorder is qualified against roughly fifteen of them.

Not every section means a full unit in a chamber. The standard explicitly allows some environments to be demonstrated on representative specimens — fluids susceptibility on material coupons rather than the assembled box, fungus resistance on a materials declaration — and others to be covered by analysis or by similarity with an already-qualified configuration. Using those provisions is not cutting corners; it is reading the standard properly, and it saves both hardware and weeks of schedule.

One section is routinely misread: the crash-safety shock of Section 7. It exists so that an installed box does not become a projectile in a survivable accident — it says nothing about the data inside surviving. Crash protection in the recorder sense is a completely separate standard and a completely different test sequence.

Reading across the two standards

The mapping is a guide, not an equivalence. Passing Method 514 does not grant you Section 8: the spectra, durations and functional criteria differ, and a certification authority will not accept a substitution without an analysis. Where the correspondence is genuinely useful is at design time — one thermal and mechanical design can be built to satisfy both, provided you know that from the start rather than discovering it after the first campaign.

Inside the lab: a typical sequence

1. Agree the test plan: sections and categories, equipment configuration, what is monitored, and above all the pass/fail criterion for each test.
2. Establish the functional baseline. The unit is characterised in a known-good state, and every later check is compared against that reference.
3. Climatic campaign: temperature and altitude, temperature variation, humidity — days in a chamber, powered and monitored, with functional checks at the extremes and after recovery.
4. Mechanical campaign: random and sine vibration on a shaker in three axes, then operational and crash-safety shocks.
5. Special environments as required by the installation: waterproofness, sand and dust, salt fog, icing, fluids, explosive atmosphere.
6. EMC campaign in a shielded or anechoic chamber: emissions first, then susceptibility, including lightning-induced transients and electrostatic discharge.
7. Final functional test, then the report — every deviation, every retest, and the environmental qualification form that will be quoted for the next twenty years.

EMC deserves its own paragraph

Electromagnetic compatibility is where qualification campaigns most often slip. It occupies nine of the twenty-six sections — power input and voltage spikes, conducted and induced susceptibility, radiated emissions and susceptibility, lightning-induced transients, direct effects and electrostatic discharge — and it is the family where the installation matters as much as the box.

In an anechoic chamber, harness routing, shield bonding and the position of the grounding point routinely decide pass or fail. Which is good news at design time: the fixes are cheap if they are anticipated — filtering at the connector, a clean chassis reference, controlled shield terminations — and expensive if they are discovered in the chamber, since each iteration costs a chamber slot.

How to read a qualification claim

• Which edition? DO-160 revisions change levels and add sections; the edition is part of the claim.
• Which sections, and which categories in each? A form, not a sentence.
• On which configuration? Hardware revision, options fitted, firmware version.
• Tested, analysed, or similar? All three are legitimate; conflating them is not.
• What was the functional criterion, and is there a report reference you can be given?

"Designed to DO-160" is a design intent. "Qualified to DO-160G, sections and categories as per the attached form, report reference X" is an engineering statement. The gap between the two is where programme risk hides — and the honest answer, when a section was covered by analysis rather than test, is simply to say so.

What it costs, and how to de-risk it

A full campaign is weeks of laboratory time and several specimens, some of which do not come back: a crash-safety shock or a flammability test consumes hardware. The real cost, though, is not the lab — it is the retest. Failing a vibration axis late in the campaign can mean a mechanical modification, a new specimen and a restart, with the programme milestone already committed.

Which is why the cheap de-risking is always early: thermal simulation before the first prototype, a shaker pre-test on a structural mock-up, a pre-scan for radiated emissions on the bench, and a functional criterion agreed with the customer before anything enters a chamber. An in-house laboratory changes the economics of all of that — you iterate in an afternoon instead of scheduling a slot in three weeks.