A battery nail penetration test machine drives a hardened steel nail through a charged lithium-ion cell at a controlled speed, then records what happens in the seconds and hours that follow. The nail is the easy part.
Almost every inquiry we get about these rigs starts with the nail and ends somewhere else: the chamber door, the exhaust duct, the force curve, or whether the machine will still be usable after a cell vents hydrofluoric acid at 600 °C. Those details decide whether your test report holds up when a customer or a notified body asks to see the raw data.
We have been building environmental and battery abuse test equipment in Dongguan since 2010. This guide covers what the machine does, which standards actually ask for the test and which ones quietly do not, what the hardware costs in 2026, and the specifications worth writing into an RFQ before you start comparing quotes.
What a Battery Nail Penetration Test Machine Actually Does
Strip it down and the machine is four systems bolted to one frame. A clamping fixture holds the cell so it cannot move. A vertical axis, usually a servo motor on a ball screw, pushes the nail at a programmed speed. A load cell sits in line with the nail to log force against displacement. And an enclosure wraps the whole thing so that when a cell goes into thermal runaway, the event stays inside.
The PLC runs a profile: approach, penetrate, hold, retract or hold in place, then observe. A data logger writes force, displacement, cell voltage, and two to four thermocouple channels to disk at 100 Hz or better, alongside video from an insulated camera.
The test in plain terms
Charge the cell to 100% state of charge. Stabilise it at 20 ± 5 °C. Mount it so the nail enters perpendicular to the largest face, through the centre. Drive the nail at the speed your standard specifies, usually 25 ± 5 mm/s for Chinese EV cell standards, or 40 to 80 mm/s for some OEM specs. Push it through the full thickness. Leave it in. Then watch for one hour, sometimes six.
Pass criteria are blunt: no fire, no explosion. Some OEM specifications are stricter and add "no visible smoke escaping the seams" or "case temperature below 150 °C". If you sell into automotive programmes, expect the stricter version.
The five things the rig has to control
Every argument about whether a nail penetration result is valid comes down to one of these five variables. If the machine cannot hold all of them repeatably, two labs will get two different answers from the same cell design.
| Variable | Typical range we build | Why it decides the result |
|---|---|---|
| Nail diameter | φ3, φ5, φ8 mm (3 to 8 mm) | Smaller nail means higher local current density at the short. A φ3 result cannot be compared with a φ8 result. |
| Nail geometry | 45° or 60° cone tip, hardened to HRC 50+ | A blunt or bent nail changes the penetration force and the contact area. |
| Penetration speed | 0.1 to 100 mm/s, closed loop ±2% | Faster penetration produces more adiabatic heating at the tip. It is not a cosmetic setting. |
| Force capacity | 20 kN for cells, 100 to 200 kN for modules and packs | Undersized frames deflect, and deflection shows up as a false force plateau. |
| Ambient control | Ambient to 80 °C heated fixture, or the whole rig inside a chamber | Cell temperature at the start of the test shifts the onset of thermal runaway by tens of degrees. |
One limit worth stating up front: a single-cell nail test tells you how one cell behaves under one mechanical insult. It does not predict how a 96-cell pack propagates. That is a different test, with a different rig and a much bigger chamber. Buyers who expect a cell-level machine to answer pack-level questions end up disappointed, and it is better to hear that before the purchase order than after.
Which Standards Require Nail Penetration (and Which Do Not)
Roughly half the RFQs we receive cite the wrong standard. A buyer asks for a "UN 38.3 nail penetration chamber", and the honest answer is that no such clause exists. UN 38.3 covers transport, and its mechanical abuse test, T6, is a crush at 13 kN. There is no nail in it.
This matters beyond pedantry. If you buy a machine configured for one standard and your customer's auditor asks for another, you have a capital asset that cannot produce the report you need. It is worth ten minutes to check the clause list before you sign.
Standards that do call for the test
GB/T 31485-2015 (China, EV traction cells): φ5 to φ8 mm steel nail, 25 ± 5 mm/s, full penetration, observe one hour, no fire and no explosion. Still written into a lot of supplier contracts even though GB 38031-2020 superseded it at pack level.
SAE J2464 (US, EV battery abuse testing): includes a penetration test among its mechanical abuse procedures.
IEC 62619 (industrial and energy storage cells): includes an internal short-circuit test, which most labs run with a nail or an induced short-circuit device.
Customer and OEM specifications: Japanese, Korean and European vehicle programmes frequently define their own nail diameter, speed and observation window. These override the published standard, so ask for the clause text, not the standard number.
Standards that do not, despite what supplier pages claim
| Standard | Scope | Nail test? | What it actually requires |
|---|---|---|---|
| UN 38.3 | Transport of lithium cells and batteries | No | T1 to T8. Mechanical abuse is T6, a 13 kN crush plus an impact drop for cells. |
| IEC 62133-2:2017 | Portable sealed secondary cells | No | Crush, thermal abuse, external short circuit, free fall, overcharge. |
| UL 1642 | Lithium cells | No | Crush at 13 kN, impact, shock, vibration, heating, projectile. |
| UL 2054 | Lithium battery packs | No | Crush, impact, shock, vibration, abnormal charging, forced discharge. |
| IEC 62619:2017 | Industrial and ESS cells | Yes, as internal short circuit | Internal short-circuit test on the cell, plus impact, drop, overcharge, thermal abuse. |
| GB/T 31485-2015 | EV traction cells (legacy) | Yes | φ5 to φ8 mm nail, 25 ± 5 mm/s, full penetration. |
| GB 38031-2020 | EV traction battery packs | No | Thermal propagation is the headline requirement, not nail penetration. |
| SAE J2464 | EV battery abuse | Yes | Penetration among the mechanical abuse procedures. |
A question that saves money: ask any supplier quoting a nail penetration machine to name the standard and the clause the configuration is built to. If the answer is "UN 38.3", ask again. The machine may still be right for you, but the paperwork will not be.
The practical upshot: if you are testing EV cells for a Chinese or Asian supply chain, you probably need a nail rig. If you are shipping consumer cells and your only obligation is UN 38.3, IEC 62133-2 or UL 1642, you need a crush rig, and the nail axis is optional. Plenty of buyers own both functions in one frame for exactly this reason.
Crush and Nail in One Chamber: When It Makes Sense
Combining the two functions shares the frame, the chamber, the exhaust train and the data system. Two separate fixtures swap onto the same platen: a flat platen for crush, a nail holder for penetration. In our experience the combined configuration runs 30 to 40 percent below the cost of two standalone machines, and it takes about 10 minutes to change over.
The trade-offs are real, though.
Contamination. After a cell vents, everything in the chamber is coated. With two functions you are cleaning one box instead of two, which is cheaper, but a dirty chamber is also the most common reason a second test in the same week produces odd numbers.
Force path. Crush needs a wide platen and a stiff frame at 13 kN and up. Nail needs a slim axis with a clean force reading. Sharing hardware means accepting a compromise on stiffness or spending more on the frame.
Throughput. If you run 20 tests a week, changeover becomes a bottleneck. If you run two a month, it does not.
Our rule of thumb for customers: under about 8 tests a month, combine. Above 15, split. In between, it depends on how many operators you have.
How Much Does a Battery Nail Penetration Test Machine Cost in 2026?
Prices vary enormously because "nail penetration machine" describes anything from a bench rig on a steel plate to a walk-in chamber with gas scrubbing and an automated fire suppression train. The ranges below are FOB China in 2026 for standard configurations, excluding freight, installation and local duties.
| Configuration | FOB range (USD) | Who buys it |
|---|---|---|
| Cell level, open bench, no enclosure, manual door | 6,000 to 12,000 | Universities, small cell assemblers, R&D benches where the cell is small and the risk is contained elsewhere |
| Cell level with explosion-proof chamber, exhaust port, camera, data system | 18,000 to 45,000 | The most common configuration. Most consumer and e-bike cell makers sit here. |
| Module level, 100 to 200 kN, chamber with relief panel and suppression | 65,000 to 180,000 | EV pack integrators, ESS module suppliers |
| Multi-station automated, carousel or twin-cell, full gas treatment | 200,000 and up | Third-party test labs running certification volume |
What actually moves the price
The nail is the cheapest component on the machine. The chamber is usually 40 to 60 percent of the total, and within the chamber the cost sits in three places:
Volume and relief design. An explosion relief panel sized for a 100 Ah cell costs far more than one sized for an 18650. Getting the relief area wrong is also the most common safety finding in factory audits.
Gas treatment. A vent port is cheap. A scrubber that handles hydrogen fluoride is not, and it is the item buyers most often delete to hit a budget.
Data and traceability. A basic logger and a validated system with audit trail, user levels and exported PDF reports are two different products at two different prices.
Two smaller levers are worth mentioning. CE marking, with the technical file and declaration, adds a few thousand dollars and several weeks. And FAT scope changes the number more than people expect: a one-day witnessed run at the factory is not the same as a five-day protocol with a third-party inspector.
How to Specify the Machine: Eight Lines for Your RFQ
Most quotes come back incomparable because the RFQ said "nail penetration tester, please quote". These eight lines will get you responses you can actually put side by side.
Cell format and energy. Prismatic, cylindrical or pouch, plus dimensions and capacity in Ah. A rig built for a 20 Ah prismatic cell is not the same machine as one for a 100 Ah cell, even at the same force rating.
Standard and clause. Name it. GB/T 31485, IEC 62619, an OEM spec, or an internal procedure. Attach the page if you have it.
Nail specification. Diameter, tip angle, material, and whether the nail must be replaced after every test. It should be.
Speed range and accuracy. Give the test speed and the tolerance you need to defend, for example 25 ± 5 mm/s with ±2% closed-loop control.
Force capacity and resolution. State the maximum force and the smallest force you need to resolve. A 200 kN cell rated to 1% cannot see a 200 N event clearly.
Enclosure requirements. Chamber volume, relief area, whether the door needs an interlock, whether you need an internal light and camera, and where the exhaust duct will go.
Data outputs. Channels, sampling rate, file format, and whether the system must export a report your customer will accept. Ask for a sample report file before you order, not after.
Acceptance terms. FAT protocol, whether you attend, calibration certificates, spare parts list, and the documentation package.
Add one more line that costs nothing: ask what the supplier's own factory uses to verify the machine before shipment. A supplier who cannot describe that test in detail has probably never run one.
Safety Design: The Part That Gets Cut First
Nail penetration is the one battery test where the failure mode is a fire inside a closed steel box. The safety hardware is not decorative, and it is also the first line item to disappear when a buyer pushes for a lower price.
Explosion relief and the pressure path
A venting cell releases gas fast. If the enclosure cannot relieve pressure, the door becomes the relief path. Look for a dedicated relief panel, a door with a mechanical latch rather than a magnetic catch, and a pressure path that leads away from where the operator stands. Ask for the relief area calculation. It should exist as a document.
Vent gas and hydrogen fluoride
Lithium-ion vent gas contains carbon monoxide, hydrogen, and hydrogen fluoride, which forms when fluorine-containing electrolyte meets moisture. HF is the one that gets underestimated. A chamber with an exhaust port and a duct to the roof is the minimum. A scrubber is the correct answer if the machine sits in an occupied building. Budget for the ducting as part of the project, not as an afterthought, because it often costs more than the optional scrubber itself.
Suppression and interlocks
Water mist or a Class D style agent for the cell, never a standard ABC extinguisher. Door interlock tied to the test state, so the chamber cannot be opened mid-run. An emergency stop that also kills the exhaust fan is a mistake we have seen; the fan should keep running. And a thermal camera or IR window is worth having if you test more than a few cells a month.
What to Check Before You Accept Delivery
Factory acceptance testing is where you find out whether you bought a machine or a drawing. These checks take half a day and catch most of the problems that would otherwise surface in your first month of production.
Speed verification. Program 25 mm/s, measure travel with a dial indicator and a stopwatch, and compare. Do it at three speeds across the range, not one.
Force calibration. Load the axis with a proving ring or a calibrated load cell at 20%, 50% and 100% of capacity. Ask for the certificate.
Chamber integrity. Confirm the relief panel is fitted, the door interlock stops the test, and the exhaust fan runs against a closed damper.
Data run on a dummy cell. Ask for a full force-displacement curve from a non-reactive sample. You are checking the shape of the curve and the sampling rate, not the numbers.
Documentation package. Wiring diagram, PLC program backup, calibration certificates, CE declaration if applicable, spare parts list with part numbers.
Spare parts. Nails, seals for the door, one thermocouple set, and the fuse and relay set. Ask what the consumables cost per year, then check that answer against the quote.
If you cannot travel, ask for the FAT to be recorded on video with a timestamp, and ask a third-party inspector to witness it. We have done both for customers in Europe and South America, and it is a normal request rather than an insult.
Battery Nail Penetration Test Machine FAQ
What is the difference between a nail penetration test and a crush test?
A nail test creates a localised short circuit by driving a small-diameter steel rod through the cell, which concentrates current in a narrow path. A crush test applies a broad mechanical load, 13 kN under UN 38.3 T6, and measures whether the cell survives deformation. Different failure mechanisms, different fixtures, different standards. Most labs eventually need both.
What nail diameter should I use?
Use whatever your standard names. GB/T 31485 calls for φ5 to φ8 mm. Small-format consumer cells are often tested with φ3 mm. If your customer's specification is silent, φ5 mm at 25 mm/s is the most widely understood starting point, but document the choice in the report.
Can one machine test cells and modules?
Sometimes, but check the force and the chamber volume rather than the marketing description. A rig rated at 20 kN for cells will not crush a module that needs 100 kN, and a chamber sized for a single 18650 will not contain a module venting event. If you need both, specify both up front. Retrofitting force capacity later means a new frame.
How long is delivery?
Standard cell-level configurations run 30 to 45 days from deposit. Module and pack level machines with custom chambers run 60 to 90 days. Add two to three weeks for CE documentation and any witnessed FAT. Freight to Europe or the US by sea adds another 30 to 45 days, so plan backwards from your audit date, not from your purchase order date.
Do I need a fume hood as well?
No, if the chamber has proper exhaust and the duct terminates outside the building. You do need to plan the duct route before the machine arrives, because a 150 mm duct cannot turn a sharp corner through a finished ceiling. If the machine goes into a shared lab rather than a dedicated test room, add the scrubber.
What is the minimum order quantity?
One unit. These are not consumer products. For private label or OEM builds with your branding, the practical minimum is three to five units, because of the tooling and documentation work involved.
Can the test run at elevated temperature?
Yes, two ways. A heated fixture holds the cell platen at up to 80 °C, which covers most specifications that call for a warm start. For a full conditioning profile, the entire rig goes inside a temperature chamber, which is a larger and more expensive build. Tell the supplier which one you mean, because the words "high temperature nail test" get used for both.
Need a Nail Penetration Chamber Built to Your Standard?
We design and build battery abuse test equipment in our own factory in Dongguan: eight product lines, our own sheet metal shop and assembly floor, and sixteen years of shipping test rigs to labs, cell makers and research institutes outside China.
Send us the standard clause you are testing to, plus your cell format and capacity. We will come back with a configuration, a force and speed specification, a chamber drawing and a price, usually within 48 hours. If a standard nail rig will not do the job, we will say so.
Custom fixtures and force ranges built to your cell drawing, not adapted from a catalogue
Witnessed FAT, calibration certificates, and a spare parts list before the machine leaves
Remote commissioning, operator training and lifetime technical support
