When a swollen battery cracks the back cover, most of the time it is not a cell quality problem — the structure simply never gave the battery room to expand. After a few hundred cycles, a lithium cell growing 3% to 8% in thickness is normal, and pouch cells are worse. In the projects we have run out of Dongguan, whenever a back cover came back cracked, the investigation landed in the same place about 80% of the time: the Z-axis clearance of the battery compartment had been designed to the dimensions of a brand-new cell. That is zero clearance.
The right move is to reserve an extra 0.5 to 1.0 mm on top of the new-cell thickness, and to make sure the compartment wall has somewhere to yield — instead of letting the back cover take the load.
The direct answer: reserve 5% to 8% of the cell's nominal thickness, never less than 0.5 mm. Take the upper bound for pouch cells, the lower bound for cylindrical and prismatic aluminum-case cells. The space belongs on the side away from the visible surface; the side that presses the cell in place must be a structural part, not the back cover. After assembly the cell should be able to move about 0.2 mm freely, with no preload.
Read the crack first, then decide where to look
When a machine comes in with a cracked back cover, I look at the crack before I look at the battery. Where it cracked basically tells you which class of problem you have.
- A radial or single coarse crack in the middle of the cell's projected area, with whitened stretched edges: insufficient Z-axis space — the cell is pushing the cover outward.
- Cracks at the four corners of the cell, initiating from a screw boss or snap root: the compartment wall constrains too rigidly, so lateral expansion has nowhere to go and stress concentrates at the fixing points.
- A crack running along the long edge of the cover, near the parting line: the cover wall itself is too thin, and cell expansion stacks on top of it.
These three have different fixes. The first is a clearance change, the second is a change to how the wall constrains, the third means changing cover wall thickness or ribs. Adding clearance without distinguishing the case can simply move the problem from the centre to the corners.
How to tell them apart: if the bulge is most obvious at the middle of the cell with both ends normal, the problem is Z-axis clearance. If the cell bulges into an arc overall and presses on the wall at all four corners, the compartment itself is undersized. If the cell looks fine but the cover has already deformed, go look at cover wall thickness and injection-molding residual stress — it is not a battery compartment problem.
Which side the clearance sits on matters more than how much you leave
We learned this the hard way on a blood pressure monitor. The machine was tight on space and the compartment clamped the cell between an upper and lower cover. To stop the cell rattling, the upper cover got a ring of ribs on the inside, and after assembly the cell was clamped solid with only 0.15 mm of clearance. It passed testing. Six months later, reports started coming in about bulging covers and degraded button feel. Teardown showed the cell had grown about 0.6 mm, and the ribs passed that force straight into the back cover.
It took three revisions to settle.
Revision 1 simply increased clearance from 0.15 mm to 0.8 mm. The cell stopped pushing the cover, but now it rattled audibly when the machine moved.
Revision 2 added 1 mm foam behind the cell. The noise went away, but the foam's compression rebound was still pushing on the back cover.
Revision 3 finally got it: the clearance has to sit on the visible-surface side, and the side against the internal structure has to be hard and fixed. The cell was located against the PCB bracket, a 1 mm cavity was left toward the back cover, and the foam was moved to the side edges to stop lateral movement. Only then was the problem gone.
Concretely, the cell's three axes need separate treatment:
- Thickness (usually Z, facing the back cover): leave a 0.5–1.0 mm cavity. Put no locating feature on this side at all.
- Width and length: leave 0.2–0.3 mm assembly clearance, using foam or a silicone strip for one-way compression, with compression ratio kept under 30%.
- The lead side: at least 2 mm. When the cell expands, the tab position shifts, and a taut wire is how you get failures.
In one sentence: expansion space belongs in the direction that carries no load, so the cell can grow, while locating is handed to the internal structural parts. Putting the clearance on the back-cover side and locating against a bracket is far more effective than simply enlarging the gap.
How to dimension it so the factory can actually build that space
The most common failure with clearance on a drawing is that the tooling shop builds to the nominal value and assembly eats the margin. A few practices we settled on:
Do not dimension the compartment with a one-sided tolerance. If the cell is 8 mm thick and the compartment is 8.5 mm deep, dimensioning it as 8.5 (+0.1/0) means molding tolerance goes upward, and adding the cell's own thickness tolerance — pouch cells are typically ±0.2 mm, which the cell maker's datasheet will state — can leave only 0.2 mm of real clearance. We now dimension the depth as 8.5 (+0.15/−0.05) and additionally require incoming cell thickness to be capped at 8.1 mm.
Do not add reinforcing ribs on the cover above the cell. Some structural engineers add cross or honeycomb ribs above the battery to stiffen the cover, and those ribs end up sitting exactly on the highest point of the bulge. If you need stiffness, put ribs around the outside of the battery area, or thicken the wall near the parting line.
One more easily missed item: screw boss height. If the bosses on either side of the compartment are exactly as tall as cell thickness plus cover wall, tightening the screws pushes the cover toward the cell. Either keep boss height below the cell surface, or move the bosses out of the cell's projected area.
For verification you do not need to wait for a real bulge. We use two methods. First, a swelling simulation block made of bakelite or 3D printed, at nominal thickness plus 0.8 mm. Fit it, tighten down, and check whether the cover deforms — then run a flashlight along the parting line to look for light leaks. Second, 70°C high-temperature aging: a pouch cell shows a measurable thickness increase at that temperature, which surfaces the problem early. The simulation block costs a few tens of yuan of material, far cheaper than discovering it in mass production.
One thing needs saying clearly: more expansion space is not better. Leave too much and the cell rattles; in a drop test the impact is then carried entirely by the solder joints and leads, and the probability of a broken connection rises markedly. Clearance and locating have to be done together — leaving space without positioning just moves the problem from the back cover to the electrical connection.
FAQ
How much expansion space should a battery compartment have?
5% to 8% of the cell's nominal thickness, with an absolute floor of 0.5 mm. Take the upper bound for pouch cells and the lower bound for cylindrical and prismatic aluminum-case cells; the final number also has to subtract the cell's own thickness tolerance and the molded part's dimensional tolerance. For an 8 mm pouch cell that means 0.5–0.65 mm. If the incoming cell tolerance is ±0.2 mm and the compartment depth molding tolerance is ±0.15 mm, the design value has to go higher still, or the clearance is consumed in the worst case. Width and length do not need this much — 0.2–0.3 mm assembly clearance is enough.
How do I tell whether a cracked back cover was caused by battery swelling?
After teardown, check whether the cell is visibly swollen and compare the crack position against the cell's projected area. If the crack sits directly above the cell with whitened stretched edges, swelling is almost certainly the cause. Conversely, if the cell is flat and within thickness spec but the cover still deformed, look at injection-molding residual stress and cover wall thickness. One useful secondary test: remove the cover and lay it flat — obvious spring-back means elastic deformation and a long-term loading problem; no spring-back means plastic deformation or cracking has occurred and the structure itself was not strong enough.
What if the clearance is large enough that the cell rattles?
Leave a cavity in the expansion direction and locate in the others. Usually that means foam or a silicone strip on the cell's width or length direction, with compression under 30%, so the cell can only expand into the reserved cavity while staying stable elsewhere. Do not put foam on the face toward the back cover: compressed foam there produces a rebound force that continuously pushes the cover, effectively applying the expansion force in advance. The correct locations are the cell's side edges and the back face against the internal bracket, with the back-cover direction left completely empty.
Can reinforcing ribs go above the battery?
Not advisable. Directly above the battery is where expansion is most pronounced, and ribs transmit force straight to the visible surface of the cover — actively reducing clearance. To stiffen the cover, place ribs around the periphery of the battery area or near the parting line. We once saw a product with a ring of ribs above the battery for compression resistance, where the rib crest sat only 0.3 mm from the cell. Slight swelling pressed against the ribs and the force transferred through them to the outer surface, leaving a visible raised ring. Moving the ribs to the periphery only, with the centre kept flat, resolved it.
Top comments (1)
tr.ee/dev-to