Mechanical performance and safety / 2026-11-22
How separator puncture strength relates to battery short-circuit risk: breaking the path into segments
How separator puncture strength affects the short-circuit rate cannot be answered with puncture strength alone. This article splits the path into mechanical blocking, dimensional retention and interface transport, and gives the criterion and test condition for each: nominal puncture strength under GB/T 36363-2018 with the retention after heat treatment (40-60% at 150 °C / 0.5 h, >20% at 180 °C / 0.5 h), thermal shrinkage 1-2% at 180 °C / 30 min, peel strength >30 N/m, and porosity about 35-45%. It also states why no conversion coefficient between puncture strength and short-circuit rate is given, and ends with a checklist that can be worked through item by item.
The note
Figures are quoted with their test condition and standard; project-specific data is confirmed one to one at the sample stage.
What this article does not provide
A conversion from puncture strength to internal short-circuit rate would require a controlled dataset: the same cell architecture with only the separator puncture strength varied, and the resulting rate of internal shorts counted. The company has not run that comparison and has not published such data.
So this article does not state how much a given increase in puncture strength lowers the short-circuit rate. Any such coefficient would be invented. What this article can do is split the path from puncture strength to short circuit into segments, and say which data and which conditions decide each segment. That is more useful than an unverifiable number: it lets you list what has to be measured in your own cell architecture.
The path has three segments, and puncture strength only covers one of them
The first is mechanical blocking. While a lithium dendrite grows, if the puncture strength of the separator is not enough to resist it, the dendrite can grow through the separator to the counter electrode and form an internal short path. This is the segment where puncture strength acts directly, and the only one where the intuition that higher puncture strength makes penetration harder holds.
The second is dimensional retention. Even with enough mechanical strength, if the separator shrinks noticeably under high temperature or abnormal duty, the contact between separator and electrode changes, and so do the conditions for dendrite growth. This segment corresponds to thermal shrinkage, not puncture strength. The published nominal figure is thermal shrinkage 1-2% at 180 °C / 30 min (GB/T 36363-2018).
The third is interface and transport. The state of adhesion between separator and electrode, and the wetting and ion transport conditions of the separator, affect how dendrites nucleate and grow at the interface. This segment corresponds to peel strength (internal criterion >30 N/m, GB/T 7122-96) and to porosity and air permeability, not to puncture strength.
Read together the three segments give the complete description. Looking at puncture strength alone means looking at the first segment, and then using a local conclusion to explain the overall behaviour.
Segment one: how puncture strength is measured, and what survives heat treatment
Puncture strength is a conditional number: specimen dimensions, loading speed and test temperature all affect the result. The company pages give the nominal value per model under GB/T 36363-2018 in gf, for example ≥400 gf for JTF-R20LA8 and ≥450 gf for JTF-R10DH. Nominal values across models are not directly comparable for a given purpose; they state the nominal band of each model.
What matters more is how much survives heat treatment. The test report states that after heat treatment at 145 °C / 1 h and 140 °C / 1 h the puncture strength of the coated separator is unchanged or slightly higher, that temperatures below 145 °C do not reduce it, and that in the 150 °C / 0.5 h window the double-side coated scheme retains 40-60% of puncture strength while in the 180 °C / 0.5 h window it still retains more than 20%.
These two sets of figures answer different questions. The absolute value answers how much there is now; the retention answers how much survives the treatment. If the thermal risk of a cell comes from a sustained extreme duty, retention is the figure to quote. If the concern is mechanical margin during assembly and normal cycling, the absolute value is more direct. State which part of the duty is being discussed before quoting either.
Segment two: thermal shrinkage only counts with its condition
Thermal shrinkage without its condition is meaningless: the same 1-2% means quite different things at 180 °C / 30 min than at 150 °C / 0.5 h. The published nominal figure is stated with 180 °C / 30 min, air atmosphere, MD and TD shrinkage, under GB/T 36363-2018.
The registered shrinkage conditions also differ between models. JTF-R20LA8 is registered at 130 °C / 1 h with MD and TD both ≤3.0%, while JTF-R10LP is registered at two conditions, 150 °C / 0.5 h and 180 °C / 0.5 h. When discussing a model, use the condition registered for that model rather than borrowing another one.
Segment three: adhesion and wetting read separately
Adhesion and wetting are independent problems. The criterion on the adhesion side is peel strength, with an internal criterion of >30 N/m under GB/T 7122-96; the criteria on the wetting side are porosity and air permeability, with separator porosity about 35-45% and air permeability about 226-366 s. Neither proves the other. High porosity and high permeability help wetting and ion transport, but the gaps they create also reduce mechanical conformity against the electrode; raising adhesion usually means a denser surface, which makes wetting harder.
On the company product these hold at the same time, and they are coexisting conclusions rather than alternatives: puncture strength retention reaches 40-60% in the 150 °C / 0.5 h window and still holds more than 20% in the 180 °C / 0.5 h window, peel strength meets the >30 N/m criterion, and separator porosity as a whole stays at about 35-45%.
Turning this into a checklist for your own cell architecture
Work through the steps in order, and attach a condition to each one.
First, write down the temperature window the cell operates in, and whether that window is a short duty or a sustained one. This decides whether the absolute value or the retention is the figure to quote.
Second, identify which shrinkage condition corresponds to that window, and which condition the model in question is registered at. If the two do not line up, the comparison is meaningless.
Third, look at the mechanical retention after heat treatment rather than the room-temperature nominal value.
Fourth, treat adhesion and wetting as two separate questions: whether the peel strength meets your own acceptance criterion, and whether porosity and air permeability fall inside the band your rate design requires.
Fifth, ask for the distribution between batches rather than the peak of a single batch. Whether a long-term cooperation can work is decided at that point.
What the company does, and where the boundary is
At the sample stage the company provides test data and failure-analysis suggestions to narrow the parameter range, and keeps a technical contact window open during mass production to handle batch variation and failure analysis. The coating composition and process parameters are trade secrets and are not published; they can be provided in a one-to-one technical discussion.
The boundary is worth stating: the breakdown above describes an order of judgement, not a prediction of short-circuit probability. How a specific cell architecture behaves against internal short circuit has to be assessed at the sample stage under that architecture’s own conditions.
Related pages
To discuss a parameter range for a specific cell architecture, contact the technical team directly.