There's a scene almost every driver knows. The chip had been there for weeks, small, harmless, almost decorative. Then one January morning, with the car still cold and the defroster on full blast, there's a brief ticking sound, and before your eyes appears a thin line running twenty centimeters in less than a second before stopping. No stone, no impact, no pothole: the car was parked.
What just happened has a precise physical explanation, and it's the same that explains why cracks grow in jumps rather than continuously, why they prefer the night and early morning, and why certain seemingly innocuous daily actions are actually their best allies. Understanding it serves two purposes: avoiding behaviors that accelerate the damage and realistically knowing how much time you can gain before intervening.
A note of honesty, because it deserves to be at the beginning and not hidden at the end: there is no method to truly stop a crack. There are ways to slow it down and prevent it from worsening. Anyone promising to stop it with a home remedy is describing something that the physics of glass does not allow.
Why Glass Behaves This Way: The Physics of the Apex
Glass is both an extraordinary and frustrating material. It is extremely hard—much harder than common steel when measuring scratch resistance—and at the same time brittle, in the technical sense: it does not deform before breaking. A metal subjected to excessive stress bends, yields gradually, and in that yielding dissipates energy. Glass does not: it holds, holds, holds, and then breaks suddenly.
This behavior has a decisive consequence when there is already a fracture in the glass. In an intact material, stress is distributed across the entire surface, and each point bears a minimal fraction. But when there's a crack, the stress no longer distributes uniformly: it concentrates at the apex of the fracture, that microscopic point where the glass is still intact and the crack is about to continue.
The amount of this concentration surprises those who discover it for the first time. At the apex of a crack, the local stress can be hundreds of times higher than the average acting on the rest of the pane. It's as if all the load the glass is bearing is channeled onto an area the size of a pinhead. And since the stress at the apex increases with the length of the crack, a self-sustaining mechanism is triggered: the longer the crack, the more easily it lengthens further.
This is why a crack behaves so differently from what intuition suggests. It doesn't grow slowly and continuously: it stays put for weeks and then advances suddenly, when the stress at the apex exceeds the threshold. Then it stops again because the geometry of the stresses has changed in the new position. It's a progression in jumps, and each jump makes the next more likely.
Where Does the Energy That Advances the Crack Come From
The concentration at the apex explains the mechanism, but something needs to provide the stress. There are four sources, and it's worth knowing them all because three of the four can be managed.
Thermal Stresses
These are by far the number one cause, which is why many people discover the crack in the morning, with the car stationary.
The mechanism is this: glass expands when heated and contracts when cooled. As long as the entire pane changes temperature together, nothing happens. The problem arises when different areas of the same glass are at different temperatures: the hot area wants to expand, the cold one restrains it, and at the boundary between the two, a real mechanical stress is generated. If the apex of a crack is in that stress field, the crack advances.
The situations that produce the worst gradient are easily recognizable and are all daily occurrences. Defrosting on full blast on a frozen windshield is the textbook case: air at forty degrees hitting the bottom part of the glass while the rest is at minus three, with a temperature difference concentrated in a few centimeters. Tepid water poured on ice is even worse because the jump is instantaneous. But the opposite case is also effective: air conditioning on full blast on a windshield heated by the sun, or the jet of cold water from a car wash on a car left in the August sun.
There's also a more insidious situation because it doesn't seem like a thermal shock: the morning sun hitting only a portion of the windshield, typically because the rest is in shadow due to a wall or another vehicle. The temperature difference builds up slowly but reaches considerable values, and the boundary zone between sun and shade moves during the morning, sweeping the glass with a moving line of stress.
Body Twists
The windshield is glued to the body and follows its movements. A car is not a rigid block: the body bends and twists continuously, and these deformations are transmitted to the glass through the bonding.
The situations that generate the greatest twists are those where the four wheels are not on the same plane: a pothole hit with one wheel, a speed bump crossed diagonally, climbing a curb with one wheel, an uneven surface. At that moment, the body twists, and the windshield undergoes a stress that concentrates, as always, at the apex of the crack.
Even parking with one wheel on the curb for an entire night is a twist, less violent but much more prolonged.
Closing the Doors
This is the most surprising cause and can be completely eliminated with a simple action. When you close a door with the car fully closed or with the windows up, the cabin air is instantly compressed and produces a pressure wave that pushes on all surfaces, including the windshield. On an intact glass, it's irrelevant. On a glass with an active crack, it's a stress that adds to all the others, repeated several times a day.
The countermeasure is simple: with an evolving crack, close the doors gently and keep a window slightly open, so the air has an escape route.
Vibrations and Time
The last source is the least spectacular and the most constant. The vibrations of normal driving provide the apex of the crack with a continuous flow of small stresses. Each is insufficient to advance the fracture, but glass has a little-known characteristic: in the presence of humidity, the apex of a crack can advance very slowly even under modest loads, in a phenomenon of subcritical growth. It's the reason why a crack can lengthen by a few millimeters in a month without anything identifiable happening.
What to Do to Buy Time
Having clarified that stopping a crack is not possible, here's what really works to slow it down—in order of actual effectiveness, not how good it sounds.
- Eliminate thermal shocks. This is the intervention with the highest result-to-effort ratio. In winter, start defrosting at a moderate temperature and gradually increase as the glass warms up; never use full blast on a frozen glass, and never pour any kind of water on the ice—the scraper is slower but doesn't cause damage. In summer, before turning the air conditioner on full blast, open the windows for a few minutes and let the hot air out: the glass cools more evenly, and so does the cabin.
- Park in the shade, or at least uniformly. A windshield entirely in the sun suffers less than one half in the sun and half in the shade because, in the second case, a gradient is created. If there's no shade, an internal sunshade helps more than it seems because it reduces heat buildup in the most exposed area.
- Drive as if you had a full glass on the dashboard. Approaching potholes and bumps calmly reduces body twists. You don't need to drive slowly everywhere: you need to avoid hitting bumps with one wheel at the speed you would normally.
- Close the doors gently and with a window slightly open. It costs nothing and eliminates a stress repeated dozens of times a day.
- Cover the damage with transparent adhesive tape. It doesn't slow propagation but keeps out water and dirt. It helps preserve the possibility of a decent intervention and, if the crack is already long, prevents the fracture from getting dirty, making the defect more visible.
- Avoid high-pressure washes in the damaged area, and in general, water jets at a temperature very different from that of the glass.
- Don't park with one wheel on the curb if you can avoid it, especially for long periods.
What Doesn't Work
There are several remedies worth debunking because someone might try them and waste precious time.
Clear nail polish on the crack: it doesn't penetrate the fracture, stays on the surface, and especially once hardened, makes a subsequent resin injection impossible. The same goes for cyanoacrylate glues and generic products: they occupy the space without doing the job.
Drilling a small hole at the apex of the crack to stop its advancement. It's a technique that indeed exists in structural engineering—it's called crack arrest and works on metal sheets because the hole "spreads" the stress concentration over a larger radius. On a windshield, it's a bad advice: drilling a laminated glass already fractured, without proper equipment and cooling, is much more likely to immediately start the crack than to stop it, and in any case, leaves a hole in the field of vision.
Waiting for it to stabilize on its own. Cracks do stop, but in a position you didn't choose. The calculation "I'll wait for it to stop and then see" isn't wrong in principle: it's wrong because the point where it stops could be on the other side of the windshield.
How Much Time Can You Realistically Gain
The practical question is always this: can I make it to the end of the month, or do I risk finding the windshield crossed tomorrow?
The answer depends on three observations you can make yourself. The first is where the apex of the crack is: if it points toward the edge of the glass, propagation is more likely and faster because the edge is the tension zone. If it points toward the center of a peripheral area, you're in a more relaxed situation. The second is how long it already is: due to the self-sustaining mechanism, a twenty-centimeter crack is much more prone to grow than a three-centimeter one. The third is the recent history: a crack that has lengthened in the last week is in an active phase and will continue; one that's been stationary for two years is in an equilibrium that could hold for a long time.
A simple and surprisingly useful way to monitor the situation is to mark the apex with a fine-tip permanent marker, placing a line on the glass where the crack ends today. It's the method technicians also use to understand if a fracture is active: if after a week the crack has passed the mark, you have the answer without having to rely on memory, which in these cases is notoriously unreliable.
The uncomfortable part must also be said: a crack that crosses the driver's field of vision poses a problem that is not just one of opportunity. Uncompromised visibility is a requirement on which both traffic rules and inspection evaluations are based, and a windshield crossed by a fracture in the primary vision area is not a situation to handle calmly. The complete picture is in the guide on driving with a damaged windshield.
Cracks That Appear Without Impact
There is a category of fractures that appear without any stone hitting the glass, and it deserves a separate discussion because the cause is different and sometimes indicates an underlying problem.
They are called stress cracks and almost always start from the edge of the windshield, advancing toward the center with a typically curved path. At the origin, there's a micro-defect of the edge—a small invisible chip produced during transport or installation—combined with a stress: a thermal shock, a twist, or a permanent stress due to the installation itself.
The latter case is the one to know: a windshield installed with the glass under tension, because the frame was deformed, because the support blocks were mispositioned, or because the adhesive bead was irregular, lives constantly under stress. It doesn't break immediately: it breaks months later when the first serious thermal shock arrives. It's the reason why a spontaneous crack appearing on a recently replaced windshield deserves to be reported to the one who performed the work.
A stress crack is recognized by three signs: it starts from the edge, has no identifiable impact point—no crater, no spall—and has a sinuous rather than straight path. It is not repairable because it lacks the cavity that resin works on.
Why Winter Is the Season of Cracks
Those who work in the glass industry have always known it and don't need statistics to confirm it: the cold months are when chips become cracks. Not because the cold is harmful in itself, but due to a combination of four factors that all present together precisely in that period.
The first is ice, and with it the morning sequence that almost everyone knows: defroster on full blast because you're in a hurry, scraper, sometimes water. Every winter morning, the windshield faces one of the most violent thermal gradients of its life, and it does so repeatedly for weeks.
The second is the jump between outside and cabin. A car parked outside at minus two degrees and then driven with the heater on for an hour presents, on the two faces of the same glass, a temperature difference that can exceed twenty-five degrees. The outer glass wants to contract, the inner one to expand, and the intermediate film is in between mediating.
The third is the state of the roads. Winter leaves potholes, and potholes mean body twists, which are transmitted to the windshield through the bonding. It's also the period when gravel and anti-ice grit spread on the asphalt produce most of the new chips: January cracks often originate from December impacts.
The fourth is less intuitive and concerns humidity. The apex of a crack in glass advances more easily in the presence of water, due to a phenomenon of stress corrosion where water molecules weaken the bonds precisely at the most stressed point. A humid winter is, from the perspective of a crack, a favorable environment.
The practical consequence is only one and it's worth making it clear: a chip discovered in autumn should be treated before winter, not after. It's the time of year when the repairability window closes more quickly.
The Table of Daily Actions
Putting it all together, this is the practical translation: what each habitual action does to an existing crack, and what to replace it with.
| Habitual Action | Effect on the Crack | Alternative |
|---|---|---|
| Defrosting on full blast on frozen glass | Violent and localized thermal gradient: the number one cause | Start lukewarm and gradually increase as the glass warms up |
| Tepid water on ice | Instant thermal shock, the worst case | Scraper, or wait with moderate defrosting |
| Air conditioner on full blast on a scorching car | Inverse but equally effective gradient | Windows open for a few minutes, then air conditioner |
| Parking with half the windshield in the sun | Slow but prolonged gradient, with moving stress line | Full shade, full sun, or sunshade covering the entire surface |
| Slamming the door with the car closed | Pressure wave repeated dozens of times a day | Gentle closing with a slightly open window |
| Pothole hit with one wheel | Body twist transmitted to the glass through bonding | Slow down on bumps, avoid hitting them diagonally |
| Pressure washer aimed at the damage | Water forced into the fracture, also worsens repairability | Low-pressure wash, damage covered with tape |
| Wheel on the curb for the night | Modest but continuous twist for hours | Parking on a flat surface |
None of these precautions stop a crack: together, however, they tangibly reduce the number of stresses that reach the apex of the fracture, and this translates into what really matters—time. Time to organize the intervention, find the right replacement, compare a quote without the pressure of a glass becoming unusable.
Frequently Asked Questions
Why did the crack appear at night, with the car in the garage?
Because night is when the glass undergoes the most prolonged cooling of the day, and because in a garage, the temperature drops unevenly. But the most common explanation is another: the crack had already started during the day and propagated gradually, and you noticed it in the morning because it's when, with the oblique light and the fogged glass, defects become visible. The fogged glass is an exceptional revealer of cracks because moisture condenses differently along the fracture.
Can the crack stop and not move anymore?
Yes, and it often happens. A crack stops when it reaches a configuration where the stress at the apex drops below the critical threshold. It can stay there for years. The problem is that the balance depends on factors that change—temperature, loads, body geometry—and no one can guarantee its future stability.
How quickly can it grow?
When it advances, a crack in glass propagates at very high speed: the run you see developing "before your eyes" happens in a fraction of a second. What is slow is the interval between one jump and the next, which can be months. This is why it doesn't make sense to try to "see" a crack as it grows: it's either stationary or has already reached where it was going.
Can I continue driving with a long crack?
Mechanically, the car functions, but there are two serious considerations. The first is visibility, which is a requirement on which traffic rules and inspections are based. The second is that the windshield is a structural element: it contributes to the rigidity of the cabin and acts as a counterplane to the passenger airbag, and an extensive fracture reduces both functions. It's not a problem during normal driving: it is at the moment when those functions would be needed.
Can the crack lead to the windshield breaking while driving?
A laminated windshield doesn't collapse and doesn't fall: even crossed by multiple cracks, it stays in place thanks to the inner film. What can happen is that the fracture branches out to the point of seriously compromising visibility, which on the highway is a situation to manage immediately by pulling over. Sudden breakage with projection of fragments into the cabin is instead the behavior of tempered glass, which is not allowed on windshields precisely for this reason.
Does it make sense to repair an already long crack?
Rarely. Resin repair works well on compact cavities and poorly on long, thin fissures, where the resin struggles to penetrate fully. Some short cracks in a favorable position are treated with the aim of stabilizing them, but the optical result is modest. The complete criteria are in the guide on repair or replace.
If I use a sunshade in summer, does the crack worsen or improve?
It improves the situation because it reduces heat buildup and thus the extent of shocks when restarting with the air conditioner. The only precaution is to use it over the entire surface and not just half: a sunshade covering half the windshield creates exactly the thermal gradient you want to avoid.
Can the windshield wiper area heater start a crack?
It can contribute, yes. The heater in the wiper rest area heats a narrow strip at the base of the windshield while the rest remains cold, and it is by definition a localized thermal gradient. It's not a reason not to use it, but if there's already damage in the lower part of the glass, it's better to activate it with the car already at temperature rather than on the frozen glass.
Does a crack affect driver assistance systems?
If it crosses the area through which the camera views the road, yes: it introduces an optical distortion in an area that must be optically clean. Some systems may signal an anomaly or deactivate; others continue to function with degraded data, which is the least reassuring scenario. The topic is explored in ADAS and recalibration.
My windshield is cracked in two different unconnected points: is it a strange case?
No, it's more common than it seems, and it usually simply means that the glass has received two impacts at different times. It becomes interesting when both cracks start from the edge without impact points: in that case, they are probably stress cracks, and it's worth asking if there's an installation stress at the origin, especially if the glass was recently replaced.
Can intense cold alone crack an intact windshield?
Uniform cold no: the glass contracts uniformly, and nothing happens. What cracks is the gradient, that is, the difference between areas. It's the reason why cars that spend the night at minus fifteen degrees don't crack on their own, while those on which tepid water is poured on the ice sometimes do.
Do I have to wait for the crack to stop growing before replacing the glass?
No, there's no technical reason to wait: the new glass is an independent piece, and the state of the old one doesn't affect it. The only situation where waiting makes sense is when the crack appeared on a recently replaced glass, and an installation stress is suspected: in that case, the issue to clarify is the preparation of the frame, because reinstalling a new glass on the same seat without correcting the problem means repeating the experience in a few months.
Why is the crack more visible when the glass is fogged?
Because along the fracture, the surface of the glass has a different micro-roughness and a slightly different temperature from the rest, and condensation deposits differently. The result is that the crack emerges very clearly on a fogged glass, while during the day, on dry glass, it can be almost invisible. It's the reason why many people discover damage in the morning and are convinced it appeared during the night, while often it had been there for days.
Does a tinting film or applied film slow down propagation?
No. A film applied to the surface has no effect on the stress acting at the apex of the crack, which is an internal phenomenon of the material. It can at most hold fragments in case of breakage, which the inner film already does on the windshield. On the windshield, moreover, the rules on applying films are the strictest precisely because they concern the primary field of vision: the general framework is in the guide on tinted windows and regulations.
If the windshield is already to be replaced, do I still need to be careful?
It makes sense to continue being careful for a practical reason: as long as the glass is intact enough to ensure visibility, the car remains fully usable, and you can organize the intervention calmly, verify the correct variant of the replacement, and compare proposals. If instead, the crack crosses the field of vision, the car becomes difficult to use, and you find yourself deciding in a hurry, which is the worst condition for making a sensible choice. The path to identify the right glass is in the guide on license plate, VIN, and Eurocode.