EMI gasket failure modes: why a gasket that shielded stops, and how to tell which mode

The modes in one table

Each row names the mode, what you see, what caused it and the check that confirms it; the sections that follow carry the data.

Galvanic corrosion at the flange

The handbook states the three conditions as a battery: two unlike metals, an electrical path between them, and an electrolyte able to dissolve the less noble one; a gasket between aluminum or steel flanges supplies the first two by nature, salt spray or condensed humidity supplies the third, and corrosion scales with the total galvanic current [S9]. MIL-STD-889D describes accelerated attack on the anodic member at a rate set by the galvanic current, and its compatibility criterion is a calculated anodic corrosion rate of at most 0.009 mil per year in seawater, with anything above requiring protection and compatibility not excluding crevice or pitting corrosion [S5]. The handbook's 168 hour salt fog coupon test on 6061-T6 aluminum gave weight losses of 2.2 mg for passivated silver-aluminum fluorosilicone, 19 and 22 mg for the other silver-aluminum grades, 37 mg for silver-glass, 167 mg for a nickel-graphite grade, 237 mg for pure silver and 281 mg for silver-copper, with nickel powder and nickel fiber at 25.5 and 10.3 mg and carbon at 35.2 mg [S9]. In the same test silver-bearing grades stayed near 7 milliohm-cm before and after while nickel-bearing grades rose from tens to hundreds of milliohm-cm, and one rose past 1000 [S9]. Newer nickel-aluminum fillers report the opposite ranking, up to twice the galvanic corrosion resistance of silver-aluminum in salt spray testing on one manufacturer's page and the best result in the 504 hour salt fog testing that led NASA to select it; this is the disputed point, because nickel powder in 2000 and nickel-plated aluminum today are different fillers, and the modern claim is limited to that filler [S32][S18][S9].

Diagnosis is white or gray corrosion product at the flange edge and under the footprint, pitting of the flange face, a rise in gasket-to-flange resistance, and a fall in shielding that the NASA protocol tracks by re-measuring shielding effectiveness at fixed frequencies after exposure [S18]. Corrosion potentials in 5 percent sodium chloride put Monel mesh at -125 mV against a saturated calomel electrode, close to silver at -25 and far from aluminum at -730, which is why Monel's corrosion resistance refers to its own oxidation and not to its compatibility with aluminum, and why it needs extensive edge sealing there [S9]. The pairing rules and the finish table are on the galvanic compatibility guide.

Flange pretreatment under salt fog and humidity

NASA's chrome-free coating study is the reference case for the flange side of the joint: hexavalent chromate pretreatment was known to lose shielding effectiveness after environmental testing, and the question was whether chrome-free Type II coatings would do better or worse [S18]. The exposures were 100 thermal cycles from 0 to 100 C, ASTM B117 salt spray for 168 and then 1008 hours, 1000 hours at 85 C and 85 percent relative humidity, and twelve months of beachfront marine exposure, with shielding from 50 MHz to 18 GHz at three frequencies per decade before and after, plus contact and surface resistance [S18]. The slides show no pitting on 5052-H32 with either the hexavalent or the chrome-free coating at six weeks of salt spray; the consolidated post-exposure shielding plots are images and their values are not quoted here [S18]. ASTM B117 itself warns that salt spray results seldom correlate with natural environments when used alone, and MIL-STD-810H Method 509.7 is the DoD counterpart [S28][S29].

Compression set and loss of contact pressure

Rubber deformed for a time keeps some of the deformation; the handbook defines compression set as the retained deformation expressed as a percentage of the initial deflection per ASTM D395 Method B, with the worked example that 30 percent set at 25 percent deflection means 92.5 percent recovery of the free height [S9][S30]. Over-compression accelerates it, and flanges that bow between bolts over-compress the gasket at the bolts [S9]. Stress relaxation is the companion effect: the load falls to about 70 to 75 percent of its initial value within an hour as the polymer rearranges [S9]. Published set values run 25 to 35 percent for most silver and nickel grades at 70 hours and 100 C, 45 to 60 percent for pure-silver and carbon grades in the November 2000 handbook, and 40 to 50 percent on the EPDM rows of one catalog [S15][S14][S11][S9]. All-metal knitted mesh takes a set, so TE advises against frequent opening; beryllium-copper fingerstock shows none within its working range but is damaged by over-compression [S34][S35].

Loss of contact pressure lowers shielding, and plane-wave attenuation is the most pressure-sensitive; for a gasket that is opened and reseated the handbook's rule is that the change in gasket thickness must not exceed twice the maximum mismatch between the mating surfaces [S9]. Under-deflection from tolerance stack-up, cover bowing and non-conformity can end in complete loss of cover-to-gasket contact over time [S9]. The signs are a gasket that no longer springs back to its groove depth, a flattened footprint at bolt locations with a slack stretch between them, shielding that fails at the lowest-pressure point between fasteners, and bolts that have lost torque in cold service or under vibration [S9]. The remedy is groove control or compression stops, torque to the relaxation factor, closer fasteners or stiffer flanges, all on the design guide [S9].

Tearing, extrusion and rupture

Conductive elastomers have low tensile and tear strength, with tensile minima of 80 to 600 psi and tear minima of 20 to 75 pounds per inch on the tables, so the handbook limits elongation to under 10 percent in service, warns that tall gaskets roll over when loaded, and states that a solid elastomer cannot change volume, so an over-filled groove damages the gasket [S9][S15]. Its worst-case figures name three damage routes: a gasket too tall for the minimum groove depth (deflection past the elastic limit, fracture), a maximum gasket in a minimum groove (over-fill, damage), and gas pressure too high for the gap (blow-out) [S9]. Waveguide gaskets creep into the opening unless reinforced with expanded metal, compression stops prevent rupture where compression is uncontrolled, and the soft 30 to 45 Shore A grades are called out for low tear strength [S9]. The signs are a split or nicked cross-section, flash extruded at the groove edge after assembly, and material missing at bolt holes on flat gaskets where the holes sat closer to an edge than the sheet thickness [S9].

Conductivity loss with time

Oxides on most metals are hard and insulating, and silver's are soft and conductive, which is why the specification has a heat-aging test and why every manufacturer table lists a resistivity ceiling after aging above the as-supplied ceiling [S9][S15]. The published pairs: 0.010 to 0.015 ohm-cm after aging for silver grades against 0.004 to 0.012 as supplied; 0.250 for nickel-graphite against 0.100; 0.200 to 0.250 for nickel-aluminum against 0.150 to 0.250, with the aging conditions per grade given as 48 hours at 100, 150 or 200 C, or 1000 hours at 125 C [S15][S14]. Fabric-over-foam surface resistivity rose from 0.009 to 0.014 ohm per square after 168 hours at 85 C, and from 0.007 to 0.017 after 168 hours at 121 C, and compression cycling raised a fabric gasket's resistance from 0.006 to 0.032 ohm over 10,000 cycles at 50 percent deflection [S9]. In salt fog the nickel-bearing grades lose conductivity and the silver-bearing grades hold, which is the handbook's stated conclusion [S9]. Silver migration under humidity and bias is documented in electronics generally but no source read gives data for it in gasket elastomers, so it is not quantified here; the handbook's silver-specific handling note is to store silver-bearing grades away from sulfur-bearing materials such as sulfur-cured neoprene and cardboard, and to clean with water or alcohol and mild soap, never aromatic or chlorinated solvents [S9].

Fluids act on the binder rather than the filler: Laird®'s fluid table rates silicone poor in DS-2 decontaminant and not for use in concentrated acids and bases, where EPDM is rated fair to good on acids and good on bases, and silicone is not for use with hydrocarbon fuels, where fluorosilicone is rated good [S11]. No manufacturer figure for the effect of cleaning agents on gasket shielding was found in the sources read, so none is quoted. The check is a four-point or two-point resistance reading against the sheet's as-supplied and aged ceilings, a look for tarnished silver or hard oxide at a Monel or aluminum contact, and a shielding loss that tracks exposure rather than assembly [S15][S9].

Installation errors

These are the errors the handbook names, most of them visible at the joint before any test.

Repeated door openings

A door or panel opened often is the case where form matters more than material. Beryllium-copper fingerstock gives a wiping contact and shows no set within its working range, but over-compression damages the fingers [S35][S17]. All-metal knitted mesh takes a set, so TE advises against frequent panel opening with it [S34]. The handbook's fabric-over-foam data show resistance rising from 0.006 to 0.032 ohm over 10,000 cycles at 50 percent deflection, a measured aging figure rather than a failure [S9]. For an elastomer gasket that is reseated, the rule is that the change in gasket thickness must not exceed twice the maximum mismatch between the mating surfaces, and the forms guide sets the seven forms side by side [S9].

Diagnosis order

The checks the handbook and datasheets describe, starting at the joint:

1. Check the joint, not the gasket: measure flange-to-gasket resistance and look for paint, chromate damage, corrosion product and oxide; the handbook ranks junction conductivity first [S9].
2. Check compression: measure free height against groove depth and the deflection window for the profile; compare bolt-adjacent and mid-span footprints; verify torque and that stops exist [S9].
3. Check the material's electrical state against the sheet: as-supplied, aged and post-vibration resistivity ceilings [S15][S14].
4. Check the exposure history: salt fog, humidity, fuel against the binder, cleaning agents, and temperature above the grade's continuous limit [S11][S15].
5. Only then re-test shielding, remembering that the fixture value never equals the flange value and that a real enclosure can only be assessed by testing it [S15][S9].

The material ceilings are on the conductive elastomers page, the type letters on the MIL-DTL-83528 guide, and the test methods on the testing guide. Send a photo of the joint with the drawing and the exposure history through the request a quote form. Name the filler or MIL-DTL-83528 type letter your drawing calls out, and the quote comes with the maker's datasheet.

Measured values on this page are the cited handbook, datasheet and NASA figures; where a number is disputed or missing the text says so.