Zinc whiskers and data center contamination control field guide

The zinc-whisker threat, and why a data center cares

Zinc whiskers are microscopic conductive crystals that grow over years out of electroplated-zinc surfaces, and in a data center the classic surface is the underside of the raised-floor tiles. They break loose, ride the underfloor air up through the perforated tiles, and land inside running equipment, where a single whisker bridged across the wrong two points causes a short. The failure that follows is the worst kind: intermittent, hard to trace, and easy to blame on the equipment.

What makes this a contamination problem and not a wiring problem is the air path. The plenum under a raised floor is a pressurized duct that feeds the cold aisles, the same path the raised-floor acceptance work and the aisle-containment scheme depend on. Whatever sheds into that plenum gets distributed to every rack the floor cools. A whisker source under one tile does not stay under one tile.

Zinc whiskers are one of three contamination threats that share that air. The other two are particulate, the dust and debris that fouls filters and heat sinks, and corrosive gases, the sulfur and chlorine compounds that eat the silver and copper on the boards. The good programs treat all three as one problem, because the same plenum, the same filters, and the same monitoring catch or miss all of them together.

What are zinc whiskers?

Zinc whiskers are thin, hair-like crystals of pure zinc that grow out of an electroplated-zinc coating over time, driven by internal stress in the plating. They are small enough that you can stand over a tile and miss them, often a fraction of a millimeter up to a few millimeters long and a few microns across, and they are conductive because they are metal.

The growth mechanism is mechanical, not chemical. When steel is electroplated with zinc and passivated, the process leaves compressive stress locked into the coating. The zinc relieves that stress by extruding filaments out of the surface, the same family of behavior seen in tin whiskers on plated electronics. The exact triggers and growth rates are still studied and debated, so treat any claimed timeline as approximate, but the field pattern is consistent: electroplated zinc grows whiskers, and older platings are the ones found shedding.

The distinction that matters is the plating method. Whisker growth is associated with electroplated zinc, the bright, thin, decorative-looking finish, far more than with hot-dip galvanizing, which lays down a thicker, differently structured coating. That is why two metal parts in the same room can behave differently, and why you cannot judge the risk by the fact that something is zinc-coated. You judge it by how the zinc got there.

Where zinc whiskers come from

The source you find first and most often is the underside of older raised-floor panels. Many access-floor tiles were finished on the plenum side with an electroplated-zinc coating, and that surface sits in the airstream for the life of the floor. It is the textbook zinc-whisker source in a data center, and it is the reason the raised-floor acceptance work and any floor reuse decision should include a look at the tile bottoms.

The tiles are not the whole story. Galvanized or zinc-plated cable trays, ladder rack, strut, threaded rod, pedestals, brackets, and assorted plated hardware all sit in or near the air path, and electroplated examples of any of them can shed. A retrofit that adds plated strut and tray under the floor can introduce a fresh zinc source into a room that was clean.

The trap is reused and salvaged material. Tiles pulled from an older hall and dropped into a new one carry their whiskers with them, and a stack of secondhand floor panels is a contamination source nobody put on the bill of materials. When in doubt about a part's plating and age, treat it as suspect until a sample says otherwise, because the cost of inspecting a tile is nothing next to the cost of a hall full of intermittent faults.

How do zinc whiskers cause equipment failures?

A whisker causes a failure when it breaks loose, travels in the air, settles inside powered equipment, and bridges two points at different potential. Because the whisker is conductive metal, that bridge is a short. Sometimes it draws enough current to vaporize itself and clear, leaving a fault that came and went with no part to replace, and sometimes it holds and takes something out.

The reason this is feared out of proportion to its size is the diagnosis. A whisker short is intermittent and location-dependent, so the symptom is a server, switch, or power supply that throws an error, gets swapped, tests fine on the bench, and then the replacement fails too because the contamination is in the room, not the box. Crews chase firmware, power quality, and bad batches for weeks before anyone lifts a floor tile and shines a light under it.

Disturbance is what turns a slow problem into a sudden one. Whiskers that grew undisturbed for years can sit in place until someone pulls tiles, runs cable, or stirs the plenum during a project, and the mechanical agitation showers loose whiskers into the air all at once. A wave of unexplained failures that starts right after underfloor work is a zinc-whisker signature until proven otherwise.

How do you detect zinc whiskers?

You confirm zinc whiskers in two stages: a visual screen in the field, then a lab analysis on a sample. The field screen is a strong flashlight held at a low, raking angle across the suspect surface, usually the underside of a lifted floor tile. Whiskers catch the light and glint like fine fuzz when the beam skims across them, where straight-on light misses them entirely. Pull a few tiles from different parts of the room and look at the plenum-side coating.

The flashlight tells you something is there. It does not prove what it is, and dust or fiber can fool the eye. To confirm, take a tape-lift sample: press a piece of clean conductive tape to the surface, lift it, and send it to a lab for microscope or scanning electron microscope (SEM) analysis. The SEM resolves the crystal morphology and, with elemental analysis, confirms the filaments are zinc rather than fiber, dust, or some other artifact. That is the positive identification a remediation decision should rest on.

Handle the sampling like evidence collection, not housekeeping. Disturb the surface as little as you can while sampling, bag and label what you pull by location, and do not start brushing or vacuuming tiles to get a look, because that is exactly the agitation that launches whiskers into the air. Confirm first, plan the response, then disturb under containment.

The underfloor airflow that spreads them

The reason a localized zinc source becomes a whole-room problem is the plenum airflow. In a downflow design the underfloor space is pressurized and the air comes up through the perforated tiles in the cold aisles to feed the racks. Anything light enough to lift off a surface in that plenum gets carried with the air, and a zinc whisker is very light. The same pressurized path that the aisle-containment scheme relies on to deliver cold air also delivers contamination.

That airflow is why the perforated tiles matter twice. They are where the cold air enters the aisle, and they are the doorway through which loose whiskers leave the plenum and reach the rack inlets. A heavily perforated, high-airflow tile in front of a dense rack is moving a lot of air, and it carries whatever is loose underneath it straight at the equipment drawing that air in.

It also explains why disturbing the plenum is so risky. Pulling tiles, dragging cable, or balancing airflow stirs the underfloor air and resuspends settled particles and whiskers, then the running fans distribute them. Any plenum work in a hall with a suspected zinc source should assume it will mobilize contamination and should plan containment and cleaning around that, not treat it as an afterthought.

Remediation: removing the source

Remediation has two halves that have to happen together: remove the zinc source and clean up what it already shed. Removing the source usually means replacing the affected floor tiles, and any plated trays, strut, or hardware found shedding, with non-whisker materials. Cleaning without removing the source buys you a few months until the surviving whiskers grow back into the air. Removing the source without cleaning leaves the population already loose in the plenum and the equipment.

The order and the containment are what separate a fix from a fresh outbreak. The single worst move is to start yanking whisker-laden tiles in a live hall with the fans running, because that aerosolizes the contamination into the air feeding the racks. A controlled remediation isolates the work area, manages or shuts the relevant airflow, removes the source material into sealed bags or wrap so it is not carried through the room, and cleans the exposed plenum and surfaces before the air path is put back in service.

Scope it from the sample evidence, not from one bad tile. Whiskers do not stay where they grew, so finding them under tile A does not mean the equipment near tile A is the only thing exposed. The lab results and the airflow map together tell you how far the contamination likely traveled, and that defines the cleaning boundary. Underscope it and you leave a reservoir that reseeds the room.

Decontamination cleaning

The cleaning side of contamination control is a specialized critical-environment service, not a janitorial one, because the goal is to capture and remove particles without redistributing them. The tools are HEPA-filtered vacuums and surface methods that trap what they lift rather than blow it around. A standard shop vac or a feather duster in a data hall makes the problem worse by exhausting fine particles back into the room and resuspending what was settled.

The work covers more than the visible floor. The underfloor plenum, the surfaces in the room, the tops and faces of racks, and where access and equipment status allow, the interior airflow paths of the equipment all hold settled contamination. Plenum cleaning is the part that matters most for zinc whiskers, because that is where the source lived and where the loose population settled. Surfaces get cleaned to keep particles from re-entering the air with foot traffic and rack work.

The discipline is in not spreading it. Work clean to dirty or in the direction the airflow and access dictate, capture at the source, and do not stir what you are trying to remove. For a zinc-whisker job this is doubly true, because every disturbed surface can release more whiskers. Pair the cleaning with the source removal and a verification check, so you can show the room came back inside its cleanliness target rather than just that it got cleaned.

Prevention: keep electroplated zinc out of the airstream

Prevention is a specification decision made before the floor and the trays go in: keep electroplated zinc out of the airstream. For floor tiles that means specifying panels finished on the plenum side with a non-whiskering coating, commonly powder-coat or an aluminum understructure rather than a bright electroplated-zinc bottom. For trays, strut, and hardware in or near the air path, it means choosing finishes that do not grow whiskers and confirming what was actually furnished against what was specified.

The cheap version of this decision is the expensive one later. A floor or a cable-management package chosen on price, with electroplated-zinc parts in the plenum, can save a little at construction and cost a remediation and a run of mysterious failures once the room is loaded. The material premium for a non-whiskering finish is small against the cost of cleaning a live hall and chasing intermittent faults.

Write it into the submittals and verify it on delivery. Call out the floor-tile underside finish and the in-plenum metal finishes in the spec, then check the actual product, because substitutions happen and a salvaged or value-engineered tile can carry the very plating you tried to exclude. The same vigilance applies to any later retrofit, since the easiest way to reintroduce zinc to a clean room is a strut-and-tray addition nobody screened.

Particulate contamination: dust, debris, and what it fouls

Particulate is the broad, everyday contamination threat, and it comes from everywhere: outside air pulled in through the cooling system, construction and fit-out debris, packaging and cardboard, and the people and carts that move through the room. Most of it is not exotic. It is dust, fiber, and fines that accumulate wherever the air slows down or the surface is cool.

What it does is mechanical and thermal. Particulate loads up filters so they have to be changed more often and, if neglected, restricts airflow. It settles on heat sinks and in fan paths, insulating components so they run hotter than the cooling design intended. Conductive or hygroscopic dust is worse, because dust that holds moisture can become conductive and bridge fine pitch, and in that state it starts to act like the gaseous and whisker threats rather than just an insulating blanket.

Fine particulate is judged by air cleanliness class and controlled by filtration, sealing, and cleaning, the topics in the sections that follow. The practical field point is that particulate is continuous. Whiskers and corrosive gas are conditions you can sometimes design out, but dust is generated every day by normal operation, so the control is an ongoing program, not a one-time fix.

What is gaseous contamination and creep corrosion?

Gaseous contamination is corrosive gas in the air that chemically attacks the metals on circuit boards, and the failure it drives is creep corrosion. The usual culprits are sulfur-bearing gases such as hydrogen sulfide (H2S) and sulfur dioxide (SO2), and chlorine compounds, drawn in from outdoor air near industry, agriculture, water treatment, paper mills, or heavy traffic. They corrode exposed copper and silver on the board, and the corrosion product creeps across the surface until it bridges adjacent features and shorts them.

The exposure that makes this current is the move toward lead-free, RoHS-compliant boards. Finishes such as immersion silver (ImAg) are reported as the most susceptible to creep corrosion, with organic solderability preservative (OSP) coatings affected to a lesser degree, so a modern board can be more vulnerable to a corrosive room than an older one was. Treat that as the documented pattern rather than a universal rule, and check the equipment manufacturer's environmental requirements for the gear you actually run.

Gaseous contamination is the threat most often missed, because there is nothing to see. A room can pass a white-glove particulate inspection and still be slowly corroding boards if it sits in a corrosive airshed and the gas-phase side was never assessed. The way you find out is reactivity monitoring, which is the next section.

What is ISA 71.04?

ANSI/ISA-71.04 is the standard that classifies how corrosive an environment is by measuring how fast it corrodes metal, expressed as severity levels G1 through GX. It is the reference the data center industry uses to put a number on gaseous contamination instead of guessing. The classification is built on reactivity monitoring: clean copper and silver coupons are exposed in the space for a set period, commonly 30 days, and the thickness of corrosion that builds on them places the environment in a severity level.

The levels run from G1, mild, where corrosion is not a meaningful factor in equipment reliability, through G2 moderate and G3 harsh, to GX severe. In the commonly cited version of the standard the copper coupon thresholds are roughly under 300 angstroms of growth in 30 days for G1, under 1000 for G2, and under 2000 for G3, with silver tracked alongside copper and GX above the G3 ceiling. Treat those figures as the standard's published bands and verify the exact numbers and the current edition against the document, because the standard has been revised and silver reactivity was added over time.

The target most data center guidance points at is G1. ASHRAE TC9.9 guidance recommends keeping copper reactivity in the G1 range, so G1 is the practical pass line for a hall. If coupons come back G2 or worse, the room has a gaseous problem to solve with filtration and source control before it shows up as corroded boards.

Reactivity monitoring

Exposing copper and silver coupons for a set period and reading the corrosion to classify the environment

Corrosion coupon

A clean strip of copper or silver placed in the space; the corrosion thickness on it sets the severity level

Angstrom

The length unit the coupon corrosion is measured in, one hundred-millionth of a centimeter

Air cleanliness and ISO 14644

On the particulate side, air cleanliness is classified by ISO 14644-1, the cleanroom standard, which rates a space by the number and size of airborne particles. Data center guidance does not ask for cleanroom-grade air. ASHRAE TC9.9 has pointed at ISO 14644-1 Class 8 as a reasonable target for a data hall, a level strict enough to protect electronics but reachable with normal data center filtration, sealing, and cleaning rather than cleanroom construction.

Class 8 is a practical line, not a hard mandate the way a fire code is. It is a recommended target, and the controlling requirement for any given project is what the owner and the equipment manufacturers specify. Some operators hold tighter, some accept looser in less sensitive spaces, and a fresh fit-out is held to a clean-build standard during construction that is separate from the operating class. Confirm the class the project actually specifies before treating any number as the bar.

The point of putting a class on the air is that it makes cleanliness measurable and verifiable. A particle count gives an acceptance test before go-live and a baseline to monitor against, so a later rise in particulate is visible as a number rather than discovered as a fouled filter or a hot rack.

Filtration: MERV and gas-phase

Filtration is the primary control that keeps particulate out, and it works on two streams: the air recirculated within the room and the outdoor air brought in for ventilation and economizer cooling. Common guidance is to filter the recirculated room air to around MERV 8 and to filter incoming outdoor air more aggressively, on the order of MERV 11 to MERV 13, because the outdoor air is where the heavier particulate and the corrosive gases enter. Confirm the filter ratings against the project's mechanical design and the cleanliness class it has to hold.

MERV filters handle particles. They do nothing for corrosive gas, which is a different control entirely. Where reactivity monitoring shows a gaseous problem, the answer is gas-phase filtration, chemical media such as activated carbon or impregnated media that adsorbs the sulfur and chlorine compounds out of the air. Gas-phase filtration is added where the airshed warrants it, typically on the outdoor-air intake and sometimes on recirculated air in a confirmed corrosive environment, and it is sized and selected to the contaminant, not bought generically.

Filters are a maintenance item, not a set-and-forget install. A loaded particulate filter restricts airflow and a spent gas-phase bed stops adsorbing, so both have a service life that the monitoring program has to track. The filter that protected the room last year is only protecting it now if someone is changing it on schedule and watching the pressure drop.

Positive pressure and the sealed envelope

Filtration only controls the air that goes through the filters. The way you stop dirty, corrosive, or humid air from leaking in around them is a sealed envelope held at a slight positive pressure relative to the surrounding spaces. Hold the data hall a touch positive and air flows out through the gaps at doors, penetrations, and the building perimeter instead of unfiltered outside air flowing in.

The envelope and the pressure work together, and a leak in one defeats the other. Open door sweeps, unsealed cable and pipe penetrations, gaps to adjacent loading docks or mechanical rooms, and a vapor barrier that was never finished all let outside air bypass the filters. In a corrosive airshed, a poorly sealed room can sit at G2 no matter how good the gas-phase filtration is, because the contaminated air is coming in through the holes, not the filters.

This is also a humidity control. The same sealed, positive envelope that keeps particulate and gas out keeps outdoor moisture from infiltrating, which matters because humidity accelerates the corrosion the gaseous threat drives. Seal the room, hold it positive, and the filters get to do the job they were sized for.

Humidity's role in corrosion and whisker growth

Humidity is the multiplier on the gaseous threat. Corrosion of copper and silver by sulfur and chlorine compounds runs faster as relative humidity rises, because moisture on the surface drives the chemistry. A room sitting in a corrosive airshed and run wet will corrode boards faster than the same room run drier, which is why the contamination and the environmental envelope are linked, not separate problems.

The operating envelope itself comes from ASHRAE TC9.9 thermal guidelines, which set the recommended and allowable temperature and humidity ranges for the equipment classes, the same envelope the aisle-containment work holds the cold aisle to. The contamination angle is that the high end of the humidity range, while fine for the equipment thermally, is the more aggressive condition for corrosion. Where reactivity is borderline, running toward the lower end of the allowable humidity is one lever to slow the corrosion.

On zinc-whisker growth specifically, hedge. Whisker growth is driven by internal stress in the plating, and the influence of humidity and temperature on the growth rate is studied and not fully settled, so do not promise that drier air stops whiskers. What is clear is the corrosion side, so control humidity for that reason and verify the equipment envelope against the manufacturer's requirements and the current ASHRAE edition.

Construction and fit-out contamination

The biggest particulate event in a room's life is its own construction. Drywall sanding, concrete cutting and curing, sawing, and general fit-out throw enormous quantities of fine dust, and that dust settles into every cavity, including the plenum and the spaces that will later carry cooling air. A room that was filthy during the build and never properly cleaned starts its operating life loaded with construction debris waiting to be stirred into the airflow.

The control is a clean-build protocol enforced while the work is happening, not a cleanup at the end. That means sealing the space off from adjacent dirty work, running temporary filtration, controlling dust at the source during cutting and sanding, keeping cardboard and packaging out of the clean areas, and cleaning progressively rather than letting debris accumulate. The mechanical systems that will serve the room should not be running and pulling construction dust into themselves and their filters before the room is clean.

The blunt rule is do not energize a dirty room. Bringing equipment up in a space still full of construction dust pulls that dust straight into the gear, fouls the new filters immediately, and bakes a contamination problem into day one. The cleanliness acceptance, the next section, is the gate that keeps that from happening, and it is far cheaper to hold the line before go-live than to decontaminate a live hall after.

Cleanliness acceptance before go-live

Cleanliness should be an acceptance item with a pass line, the same as power and cooling, checked before the equipment goes in. The common form is a white-glove or particle-count acceptance: the room is professionally cleaned after construction, then verified, either by a documented inspection of surfaces and the plenum or by an airborne particle count against the specified ISO 14644 class, or both. The point is to prove the room is clean, not to assert it.

This ties into the broader commissioning and turnover sequence, and it belongs next to the raised-floor acceptance work, because the floor and the plenum are exactly where contamination hides and where the cleanliness check has to reach. A floor accepted for level, load, and air-seal but never checked for what is coating the tile bottoms is a floor that passed one test and skipped another. Fold the zinc-whisker tile inspection into that acceptance rather than discovering it after load.

Where the room sits in a corrosive airshed, the acceptance should include a baseline reactivity reading, not just a particle count. A set of copper and silver coupons exposed before go-live tells you the environment the equipment is walking into and gives you the baseline the operating program will monitor against. Accept the room clean on both axes, particulate and gaseous, and the operating program starts from data instead of assumption.

The ongoing monitoring program

Contamination control does not end at go-live, because the threats are ongoing: dust is generated every day, the airshed changes with seasons and nearby industry, and whiskers keep growing on any source that was left in place. The operating program is three running checks: reactivity coupons for the gaseous side, particle counts for the particulate side, and periodic visual tile and plenum inspections for the zinc-whisker side.

Reactivity monitoring is the one most worth standardizing. Copper and silver coupons exposed on a schedule, commonly read on a monthly or quarterly cadence and trended over time, turn the invisible gaseous threat into a chart you can act on before boards corrode. A drift from G1 toward G2 is an early warning that the filtration or the envelope is losing the fight, and it shows up in the coupons long before it shows up in failures.

Particle counts and visual checks cover the other two. Periodic particle counts confirm the room is still holding its cleanliness class and catch a filter or sealing problem. Periodic visual inspection of representative tile bottoms and the plenum, with the raking flashlight, catches whisker growth on any source that survived or was reintroduced. Tie the readings to a record so a later failure investigation has a trend to read, not a blank.

Why high-density and AI halls move more contamination

Higher rack density moves more air, and more air moves more contamination. An AI or high-density hall pulls far more airflow per rack than a legacy room, so whatever is loose in the plenum or the room gets transported faster and in greater volume to the equipment. The same density that drives the aisle-containment and cooling design also raises the stakes on contamination, because the air-mover is bigger.

The gaseous side scales the same way. More outdoor air pulled in for economizer hours means more corrosive gas presented to the gas-phase filters and more chance for infiltration if the envelope leaks. A contamination program sized for a 5 kW-per-rack room is not automatically sized for a 30 kW-per-rack room, even in the same building.

The takeaway is that contamination control deserves the same scaling attention as power and cooling when density climbs. Reassess filtration capacity, monitoring cadence, and the cleanliness baseline when a hall is densified, rather than assuming the program that held the old load holds the new one.

The three contamination threats, side by side

It helps to hold the three threats in one frame, because they share the air but differ in source, what they damage, and how you control them. Zinc whiskers are a conductive-metal threat from electroplated zinc, controlled by source removal and material choice. Particulate is a dust threat from inside and outside, controlled by filtration and cleaning. Corrosive gas is a chemical threat from the outdoor airshed, controlled by gas-phase filtration and a sealed envelope.

The reason to frame them together is that a program that catches one can miss the others. A room can be spotless on particulate, sealed and positive, and still corroding boards from a gaseous problem the particle count never sees. A room can pass reactivity and particulate and still have zinc whiskers shedding under the floor that neither test was looking for. Cover all three or you have covered none of them with confidence.

What to document

A contamination program that is not written down cannot be defended when a hall starts throwing faults a year later. The record is what tells the investigator whether the room was clean at go-live, what the airshed looked like, and whether anything was reintroduced since. Capture the source assessment, the acceptance results, and the monitoring trend together, keyed to dates and locations.

Record the floor-tile and in-plenum metal finishes as furnished, any zinc-whisker sample results and the lab that ran them, the cleanliness acceptance result against the specified ISO 14644 class, the baseline and ongoing reactivity coupon readings with their severity level, the filter types and change records, and any remediation done with its scope and verification. Where a result missed the target, record what was done about it. The table below is the minimum spine.

Field checklist

Common mistakes

• Allowing electroplated-zinc tiles, trays, or hardware into the airstream, especially reused or salvaged tiles that carry whiskers in.
• Disturbing whisker-laden tiles in a live hall with the fans running, aerosolizing the contamination into the air feeding the racks.
• Energizing a room still full of construction dust, fouling new filters and the equipment from day one.
• Running no gaseous monitoring in a corrosive airshed, so creep corrosion eats boards with nothing to see and no coupon to warn.
• Filtering with MERV alone where reactivity shows a gas problem, when only gas-phase media removes corrosive gas.
• Leaving the envelope unsealed or at neutral pressure, so unfiltered outside air bypasses the filters through the gaps.
• Accepting a room on cleanliness alone with no particle count or reactivity baseline to defend later.
• Chasing intermittent equipment failures as bad hardware for weeks without lifting a tile to check for whiskers.
• Cleaning a whisker source without removing it, so the surviving whiskers grow back into the air within months.

Standards and references

The gaseous side is governed by ANSI/ISA-71.04, which sets the severity levels G1 through GX and the reactivity-monitoring method, copper and silver coupons read over a set exposure, that classifies the environment. The standard has been revised over its history, including the addition of silver reactivity, so confirm the severity bands and the method against the current edition rather than a remembered number.

ASHRAE Technical Committee 9.9 publishes the data center guidance that ties the threats together: the Gaseous and Particulate Contamination Guidelines for Data Centers, which recommend keeping copper reactivity in the ISA G1 range and the airborne particulate at ISO 14644-1 Class 8, and the Thermal Guidelines for Data Processing Environments, which set the temperature and humidity envelope the corrosion side rides on. Air cleanliness classes themselves come from ISO 14644. Treat the ASHRAE figures as recommended targets and confirm the project's specified class and the IT equipment manufacturer's environmental requirements, which control.

The cleaning and decontamination work is performed by critical-environment specialists to the project's cleanliness specification, and the standards above set the acceptance criteria those services are measured against. Cite the standard that controls the point, hedge the severity levels and cleanliness classes to the standard and the edition, and let the equipment manufacturer's requirements and the project specification override the rule of thumb when they are stricter.

Units, terms, and conversions

Contamination control borrows vocabulary from electronics reliability, cleanroom science, and HVAC, so the same idea reads differently across an ISA report, a particle-count certificate, and a mechanical spec. The terms below travel across the whole contamination scope.

Corrosion on a reactivity coupon is reported in angstroms of thickness over a stated exposure period, so always read the exposure with the number. Air cleanliness is given as an ISO 14644-1 class, and particulate concentration in micrograms per cubic meter. Filter performance is given as a MERV rating for particulate and by adsorption media type for gas-phase. Severity is the ISA G1 through GX scale. None of these stands alone without its method, so record the method with the result.

Zinc whisker

A conductive zinc crystal grown from electroplated zinc by internal stress, which breaks loose and shorts electronics

Creep corrosion

Corrosion product from sulfur or chlorine gases that creeps across a board and bridges adjacent features

ISA severity level

The G1 mild through GX severe classification of how corrosive an environment is, per ANSI/ISA-71.04

Reactivity coupon

A copper or silver strip exposed in the space; its corrosion thickness sets the severity level

ISO 14644 class

The cleanroom-standard rating of airborne particle cleanliness; Class 8 is a common data center target

MERV

Minimum Efficiency Reporting Value, the particulate filter rating; gas-phase media is rated separately

Gas-phase filtration

Chemical media such as activated carbon that adsorbs corrosive gases that MERV filters cannot capture

Plenum

The pressurized underfloor air path that distributes cooling air, and any contamination loose in it

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• What load rating does a data center raised floor need?
• What is the difference between hot-aisle and cold-aisle containment?
• Cast-in vs post-installed anchors: which should I use?
• What is the difference between cold and hot commissioning of a BESS?
• What is a CDU in a data center?