A case wraps the whole device in a soft shell and lets the outer material absorb the punishment. A naked product has no such shell, so drop resistance has to come from inside the structure: shed the impact energy before it reaches the screen, the mainboard or the battery. Our standard sequence is to fix the drop criteria first, then treat the corners and sidewalls as the primary load paths, and finally keep the expensive parts floating on elastic connections. Simply thickening the entire wall is the least efficient option — after several revisions I am fairly confident about that.
Write the criteria first, then talk about structure
A customer saying "it has to survive drops" is not a design input. Three things have to be pinned down: from what height, onto what surface, and what counts as a failure afterwards. Consumer electronics in bare form are usually specified for free-fall drops from 1.0 m to 1.2 m, following the free-fall test method in GB/T 2423.8. The 1.5 m or even 2 m figures quoted by case makers are mostly numbers with a case installed, so treat bare-product targets one notch lower.
The orientation set also has to be fixed. Real projects test six faces, three edges and one corner, and the corner drop is usually the first to fail. We keep the order fixed — corner first, then edges, then faces — so accumulated damage from earlier drops cannot contaminate the judgement.
Failure criteria are written at three levels: functional failure (will not power on, connections drop), cosmetic failure (cracks, paint loss, misaligned seams), and hidden damage (cracked solder joints, loose connectors, autofocus faults). The third level is the easiest to miss: the unit powers on fine right after the drop, and the combined assembly stress plus ageing only surfaces two weeks later. The more precisely the criteria are written, the more targeted the added ribs and adhesive become.
The corner is the main battlefield — that is where the money belongs
A corner drop has the smallest contact area, so the stress concentrates on a single point, and a thicker sidewall barely helps. The first move is to make the outer corner radius large enough: R2 mm minimum, and the larger the radius the longer the contact time and the lower the peak force. The complication is styling. A product with a hard, angular look may only be allowed a 1 mm radius, and then the reinforcement has to come back from the inner wall.
Inner corner radii are normally taken from 0.5 times the wall thickness, with the outer radius equal to the inner radius plus the wall thickness. Corner thickening should not run across the whole face: thicken the inner wall only over a certain height range at the corner and add one rib, and the weight penalty stays limited while the effect is far more obvious than uniform thickening. The transition in wall thickness has to be gradual, otherwise the corner comes out of the mold with a sink mark.
The real protection is letting expensive parts float. The screen, battery and mainboard should not sit directly against the inner wall of the housing: leave a buffer gap all around and use foam, TPU pads or double-sided tape for a floating connection, with compression set at 20% to 30%. On impact the housing deforms and absorbs energy first, so the acceleration peak reaching the screen comes down. Hard-soft combinations are also common: a PC hard shell with TPE overmolded a second time, where the soft material absorbs energy and the hard shell keeps stiffness — at the cost of an extra mold and an extra process step, which small-volume projects cannot justify.
Assembly details and the sums you have to do
Screw bosses crack at the base more often than anywhere else. Adding a radius at the boss base helps noticeably, and the boss wall should not be too thick — overly thick walls sink and raise stress at the base instead. A figure around 0.6 times the main wall thickness works, and triangular gussets add strength more effectively than thickening the boss wall.
Snaps should be designed around the drop direction. During a drop the two halves shift relative to each other and the snap is pried open, so a chamfered engagement face plus a back-out step works better than simply increasing engagement depth. The parting line should also not sit exactly on the corner impact line: offset it slightly and cracks are much less likely to run along the mold seam.
We have hit this one ourselves. On a handheld device housing, the first revision thickened the sidewall overall and the crack location did not move at all; moving the material to local ribs on the inside of the corners finally passed. The trade-offs are real: added material buys weight and sink risk, buffer pads buy assembly clearance and assembly labour, and two-shot overmolding buys tooling cost. I prefer to agree on a "drop budget" with the customer at the start of a project — how many extra grams and how much extra cost are acceptable — and then compare every option inside that frame, which makes the discussion much faster.
The last step is that you must tear the unit down after the drop. Assembly state, whether the battery is installed, whether the protective film is applied — all of it has to match production, or the results cannot be reproduced. No visible crack does not mean it passed: open it up and check the solder joints, the connectors and whether the snaps have whitened. That is the complete acceptance.
Questions we get most often
Q: What drop height should drop-resistance design be specified to?
A: Consumer electronics in bare form are usually specified for free-fall from 1.0 m to 1.2 m, following GB/T 2423.8. Quoted figures of 1.5 m and above normally assume a case, and cannot be used directly as a bare-product target. Drop height, floor material, the face/edge/corner sequence and failure criteria all belong in the technical requirements before design starts.
Q: How much does the corner radius change the result?
A: The radius is the best value item in drop resistance. Going from R1 mm to R2 mm or more lengthens contact time and lowers peak stress, often more than adding half a millimetre to the sidewall. When styling limits the radius, move the reinforcement to the inner wall: inner corner radius from 0.5 times wall thickness, plus local thickening and a rib inside the corner, with a gradual transition to avoid sink marks.
Q: How much buffer gap should be left around the screen and battery?
A: The gap itself is not the key number — the compression of the buffer pad is, normally 20% to 30% of pad thickness, so the housing deforms first, the pad absorbs next, and the screen takes the impact last. Foam or TPU depends on drop direction and available space. Compressed too hard the pad is effectively rigid and the screen sits on the housing; too loose and it rattles in assembly and transport.
Q: Does a product with no visible crack after the drop pass?
A: No. Hidden damage is only visible after teardown: cracked solder joints, loose connectors and whitened snaps often look normal at the time and only surface later. Sample condition must match production, and the sequence is function test first, then teardown to inspect solder joints, connectors and snaps. All three have to pass.
Originally published by the Hezi Industrial Design team, based on our industrial design project practice in Dongguan. Learn more: Industrial design services
Top comments (0)