The Hidden Science of Everyday Materials
Why ordinary objects break, bend, stick, and stretch—and why failure often waits until the last margin disappears. A field guide to flaws, time, and environment.
By TheMurrow Editorial
February 15, 2026
Key Points
- 1Recognize flaws as destiny: tiny scratches and microcracks often control real-world failure more than any single “strength” number.
- 2Track time and repetition: fatigue and polymer creep let damage accumulate invisibly for months or millions of cycles, then fail suddenly.
- 3Account for conditions: loading rate, temperature near Tg, and humidity/UV/solvents can erase safety margins without changing the applied load.
A wine glass slips, strikes the edge of a sink, and detonates into glittering sand. A phone screen survives months of pocket life, then spiderwebs from a waist-high drop that “didn’t even look that bad.” A plastic hook that held for years suddenly sags, then snaps, after one humid summer. None of this is mysterious. It just feels that way because failure often waits—quietly—until one moment when the last margin disappears.
Everyday objects rarely break because we exceeded some single, clean “strength.” They break because real materials live with flaws, time, and environment. They age. They absorb water. They collect tiny scratches. They deform slowly under constant load. And they carry microscopic cracks that can grow for months before they announce themselves all at once.
Engineers have names for these repeating mechanisms: crack growth, yielding (plastic flow), viscoelastic creep, interfacial debonding, and environment-assisted damage from humidity, UV, solvents, salt, and heat. The reassuring part is that these mechanisms are predictable. The unsettling part is that they are everywhere.
Materials don’t just have strength. They have time—and they have flaws.
— — TheMurrow Editorial
What follows is a field guide to why ordinary things fail, using a few hard numbers and a few deceptively simple ideas. If you’ve ever wondered why glass can be “strong” and still shatter, why metal bends instead of snaps, or why plastic turns treacherous near a warm car dashboard, the answers rhyme.
The tyranny of flaws: why “strength” isn’t the property you think it is
The most common misunderstanding about breakage is treating strength as destiny. Strength is the stress required to cause failure in an idealized test. Real objects aren’t idealized. They carry scratches, pores, and microscopic cracks—many introduced during manufacturing, many added by use.
That’s why fracture toughness matters so much. Fracture toughness, often reported as K_Ic, describes a material’s resistance to crack propagation—how hard it is for an existing crack to run. In the real world, cracks are the rule, not the exception, and K_Ic often decides whether damage stays cosmetic or becomes catastrophic.
Consider the spread in widely cited K_Ic values (commonly summarized in comparative tables, including on Wikipedia’s fracture toughness page):
- Soda‑lime glass: about 0.7–0.8 MPa·m^1/2
- Alumina (ceramic): about 3–5 MPa·m^1/2
- Typical aluminum alloys: about 14–28 MPa·m^1/2
Those numbers differ by orders of magnitude. A scratch on glass is not “a small defect.” It can be the starting gun.
That’s why fracture toughness matters so much. Fracture toughness, often reported as K_Ic, describes a material’s resistance to crack propagation—how hard it is for an existing crack to run. In the real world, cracks are the rule, not the exception, and K_Ic often decides whether damage stays cosmetic or becomes catastrophic.
Consider the spread in widely cited K_Ic values (commonly summarized in comparative tables, including on Wikipedia’s fracture toughness page):
- Soda‑lime glass: about 0.7–0.8 MPa·m^1/2
- Alumina (ceramic): about 3–5 MPa·m^1/2
- Typical aluminum alloys: about 14–28 MPa·m^1/2
Those numbers differ by orders of magnitude. A scratch on glass is not “a small defect.” It can be the starting gun.
0.7–0.8 MPa·m^1/2
Widely cited K_Ic range for soda‑lime glass, illustrating how little resistance it has to crack propagation once flaws exist.
3–5 MPa·m^1/2
Widely cited K_Ic range for alumina (ceramic)—tougher than glass, still far below metals in crack-growth resistance.
14–28 MPa·m^1/2
Widely cited K_Ic range for typical aluminum alloys, showing metals’ much higher ability to resist crack growth and absorb damage.
Glass: strong in one sense, fragile in the one that counts
Glass can tolerate high compressive stress, which helps explain why it performs well in windows and bottles. Yet a tiny surface flaw concentrates stress at its tip, and with K_Ic around 0.7–0.8 MPa·m^1/2, the crack doesn’t need much encouragement to propagate.
Ceramics such as alumina do better—3–5 MPa·m^1/2—but they still live closer to glass than to metals. Aluminum, by contrast, has enough toughness (14–28 MPa·m^1/2) to absorb energy and resist crack growth far more effectively.
Ceramics such as alumina do better—3–5 MPa·m^1/2—but they still live closer to glass than to metals. Aluminum, by contrast, has enough toughness (14–28 MPa·m^1/2) to absorb energy and resist crack growth far more effectively.
A scratch isn’t cosmetic when the material’s job is to stop cracks—and it can’t.
— — TheMurrow Editorial
A journalistic caution about “the number”
Even toughness isn’t a single immutable truth. Testing standards exist because specimen geometry, thickness, and validity criteria affect the result. For elastic‑plastic fracture characterization, engineers reference standards such as ASTM E1820 (J‑integral methods). For plane‑strain fracture toughness K_Ic, ASTM E399 is the canonical designation and is often cited in technical summaries.
For readers, the takeaway is simple: treat published K_Ic values as comparative, not absolute. The ranking—glass low, metals high—is robust. The precise digits depend on how the measurement is done.
For readers, the takeaway is simple: treat published K_Ic values as comparative, not absolute. The ranking—glass low, metals high—is robust. The precise digits depend on how the measurement is done.
Speed changes everything: why hitting and pulling don’t break the same way
Most people intuitively know that a sudden impact “feels different” than a slow bend. Materials agree. Many show rate effects: their measured resistance to crack growth and deformation can shift depending on how fast they are loaded.
A dissertation from Purdue University examining soda‑lime glass under dynamic loading reports measured fracture toughness values ranging roughly from 0.66 to 2.22 MPa·m^1/2, increasing with loading rate over the tested regime. The glass didn’t become ductile. It remained brittle and flaw‑sensitive. Yet the change is large enough to matter in the margins—especially when the margins are already thin.
A dissertation from Purdue University examining soda‑lime glass under dynamic loading reports measured fracture toughness values ranging roughly from 0.66 to 2.22 MPa·m^1/2, increasing with loading rate over the tested regime. The glass didn’t become ductile. It remained brittle and flaw‑sensitive. Yet the change is large enough to matter in the margins—especially when the margins are already thin.
0.66–2.22 MPa·m^1/2
Reported dynamic-loading fracture toughness range for soda‑lime glass in a Purdue dissertation, showing rate effects can shift measured toughness.
The everyday translation: the same object, two different failures
Load a material slowly and you give cracks time to extend and damage mechanisms time to accumulate. Hit it fast and the response can shift: cracks may see a different effective resistance, and the object may fail in a different pattern.
That doesn’t mean fast is safer. A high‑rate impact also introduces higher peak stresses, different stress waves, and a greater chance of initiating new cracks. What changes is not whether a brittle material becomes “tough,” but how the competition plays out between crack initiation and crack propagation.
That doesn’t mean fast is safer. A high‑rate impact also introduces higher peak stresses, different stress waves, and a greater chance of initiating new cracks. What changes is not whether a brittle material becomes “tough,” but how the competition plays out between crack initiation and crack propagation.
Why “it was fine yesterday” is a real statement
Rate effects also help explain why two seemingly similar accidents yield different outcomes. A phone screen can survive a slow squeeze in a back pocket and die on a quick corner impact. A glass shelf can hold a steady load for years and fail when someone sets down a heavy object with a sharp, brief jolt.
Failure isn’t only about how much force—it's about how fast you ask a material to take it.
— — TheMurrow Editorial
Fatigue: the slow-motion crack that makes failure look sudden
Some breakage is dramatic. Much of it is patient.
Fatigue is the process by which cracks grow under repeated loading—each cycle advancing damage a tiny amount until one day the remaining intact material can’t carry the load. The classic empirical relationship used to describe fatigue crack growth is Paris’ law (Paris–Erdogan):
\( da/dN = C(\Delta K)^m \)
In plain language: the crack growth per cycle depends on how much the stress intensity varies each cycle. Many parts survive thousands or millions of cycles and still fail “out of nowhere,” because the early crack growth is microscopic and the final break happens fast.
Fatigue is the process by which cracks grow under repeated loading—each cycle advancing damage a tiny amount until one day the remaining intact material can’t carry the load. The classic empirical relationship used to describe fatigue crack growth is Paris’ law (Paris–Erdogan):
\( da/dN = C(\Delta K)^m \)
In plain language: the crack growth per cycle depends on how much the stress intensity varies each cycle. Many parts survive thousands or millions of cycles and still fail “out of nowhere,” because the early crack growth is microscopic and the final break happens fast.
Real-world case study: the part that “always worked” until it didn’t
Think of a metal hinge on a frequently used door, a bike component under repetitive pedaling loads, or any part that flexes slightly each day. Early on, the crack is so small it changes nothing you can feel. Later, it reaches a length where each cycle increases the stress at the crack tip more sharply, accelerating growth. The last stage can be abrupt—sometimes one ordinary use after a long quiet history.
What fatigue teaches about design and maintenance
Fatigue turns everyday life into a materials experiment: repetition matters. That’s why engineers avoid sharp corners and notches in parts that cycle. Stress concentrators don’t just raise the chance of immediate fracture; they feed fatigue by creating a preferred site for cracks to start.
For readers, the practical insight is sobering: a product that has “held up fine” for years may be closer to the end of its life than you can see—especially if it experiences repeated stress, vibration, or flexing.
For readers, the practical insight is sobering: a product that has “held up fine” for years may be closer to the end of its life than you can see—especially if it experiences repeated stress, vibration, or flexing.
Bending, staying bent, and not snapping: elasticity versus plasticity
When a metal spoon bends instead of breaking, you are watching a material choose plastic deformation over fracture. The distinction between elastic and plastic behavior is the backbone of how everyday objects feel in your hands.
- Elastic deformation: reversible; the object returns to its original shape when unloaded.
- Plastic deformation: permanent; the object stays bent because internal structure has rearranged.
Metals often earn their reputation for “toughness” because plasticity can blunt a crack tip, spreading stress over a larger region and dissipating energy. Brittle materials—glass and many ceramics—lack that ability. Once a crack starts, it can run.
- Elastic deformation: reversible; the object returns to its original shape when unloaded.
- Plastic deformation: permanent; the object stays bent because internal structure has rearranged.
Metals often earn their reputation for “toughness” because plasticity can blunt a crack tip, spreading stress over a larger region and dissipating energy. Brittle materials—glass and many ceramics—lack that ability. Once a crack starts, it can run.
Ductility as a safety feature
Ductile materials offer warning. They bend, stretch, or deform before catastrophic failure. Brittle materials can fail with little visible deformation, which makes them feel treacherous even when they are adequate within design limits.
That’s not a moral judgment about materials. It’s a selection issue. A material suited to one job becomes the wrong choice in another when the consequences of failure change.
That’s not a moral judgment about materials. It’s a selection issue. A material suited to one job becomes the wrong choice in another when the consequences of failure change.
The unseen villain: stress concentration
Everyday objects are full of stress concentrators: scratches, sharp corners, drilled holes, and notches. These features amplify local stress, which can trigger yielding in ductile materials or initiate cracks in brittle ones.
That’s why so much design comes down to geometry. Rounded corners and smooth transitions aren’t aesthetic flourishes; they are stress management. The next time a cheap plastic clip snaps at a sharp internal corner, you’re seeing a basic principle: failure prefers the place where stress piles up.
That’s why so much design comes down to geometry. Rounded corners and smooth transitions aren’t aesthetic flourishes; they are stress management. The next time a cheap plastic clip snaps at a sharp internal corner, you’re seeing a basic principle: failure prefers the place where stress piles up.
Plastics and time: creep, chain motion, and the tyranny of temperature
Plastics fail in a way that frustrates anyone who expects a clean “before and after.” Under constant load, many polymers creep—they deform slowly over time, even if the load never increases. Hang a modest weight on a plastic hook and it may hold for days, then sag for months, then finally crack.
A key concept for understanding this is the glass transition temperature (Tg). Tg isn’t melting. It is a transition—especially relevant to amorphous regions—where polymer chains become more mobile. Above Tg, polymers tend to behave more rubbery and can flow or creep more readily. Below Tg, they tend to be stiffer and more glass-like.
ScienceDirect’s overview of Tg (in materials-science context) provides indicative values that show why common plastics behave so differently in heat:
- Polystyrene (PS): ~100°C
- PMMA (acrylic): ~105°C
- PET: ~70°C
Those numbers are formulation-dependent, but the hierarchy matters. PET’s Tg near 70°C makes it more vulnerable to softening and creep in hot environments that are not exotic at all—such as a car interior in summer.
A key concept for understanding this is the glass transition temperature (Tg). Tg isn’t melting. It is a transition—especially relevant to amorphous regions—where polymer chains become more mobile. Above Tg, polymers tend to behave more rubbery and can flow or creep more readily. Below Tg, they tend to be stiffer and more glass-like.
ScienceDirect’s overview of Tg (in materials-science context) provides indicative values that show why common plastics behave so differently in heat:
- Polystyrene (PS): ~100°C
- PMMA (acrylic): ~105°C
- PET: ~70°C
Those numbers are formulation-dependent, but the hierarchy matters. PET’s Tg near 70°C makes it more vulnerable to softening and creep in hot environments that are not exotic at all—such as a car interior in summer.
Real-world case study: the warm car problem
A plastic part that seems rigid in a cool room can become compliant in a hot vehicle. The load hasn’t changed; the material has. If the working temperature approaches Tg, long-term deformation accelerates. Creep becomes a design issue, not a user error.
Why identical-looking plastics can behave differently
Two clear plastics can look the same and fail differently because Tg, additives, and microstructure differ. Even the same polymer can shift behavior due to processing history. For consumers, the practical lesson is to treat heat exposure as a form of loading. A “light” load at high temperature can be more damaging than a heavier load at low temperature.
Environment as an accomplice: humidity, UV, solvents, salt, and heat cycles
People often blame “bad luck” when a product fails after a season outdoors or a year in a damp bathroom. Environment is not background noise; it is an active participant in damage.
The research lens here is straightforward: everyday failures often involve environment-assisted damage—humidity, UV, solvents, salt, and thermal cycling—acting alongside mechanical stress. The environment can:
- Promote crack growth (especially in brittle materials)
- Weaken interfaces (leading to interfacial debonding)
- Change polymer mobility (accelerating creep)
- Introduce chemical attack or stress corrosion in susceptible systems
The research lens here is straightforward: everyday failures often involve environment-assisted damage—humidity, UV, solvents, salt, and thermal cycling—acting alongside mechanical stress. The environment can:
- Promote crack growth (especially in brittle materials)
- Weaken interfaces (leading to interfacial debonding)
- Change polymer mobility (accelerating creep)
- Introduce chemical attack or stress corrosion in susceptible systems
Interfacial debonding: why composites and coatings fail “at the seam”
Many modern products rely on interfaces: coatings on substrates, adhesives between parts, fiber-reinforced composites, laminated structures. Interfaces are efficient and lightweight, but they are also places where water ingress, thermal expansion mismatch, or chemical exposure can erode adhesion.
When the interface fails, the material itself may still be strong. The structure fails because its pieces stop acting like a single unit.
When the interface fails, the material itself may still be strong. The structure fails because its pieces stop acting like a single unit.
Heat cycles: damage by repetition, not extremes
Thermal cycling—repeated heating and cooling—induces expansion and contraction. Even if the temperature range seems mild, repeated cycles can aggravate existing flaws and promote debonding. Add humidity and you have a reliable recipe for “it lasted a year and then fell apart.”
A fair perspective recognizes that consumers can’t control all environmental exposures, and manufacturers can’t make everything indestructible. Still, many frustrations trace back to predictable interactions between material choice and service environment.
A fair perspective recognizes that consumers can’t control all environmental exposures, and manufacturers can’t make everything indestructible. Still, many frustrations trace back to predictable interactions between material choice and service environment.
What to do with this knowledge: a practical failure checklist
Understanding failure mechanisms doesn’t require a lab. It requires asking better questions about what you’re seeing.
A reader’s checklist for diagnosing (and preventing) breakage
When something fails—or when you want to prevent failure—run through these factors:
- Flaws: Are there scratches, chips, or sharp notches? Brittle materials are especially sensitive.
- Time under load: Is the part holding a constant load for long periods? Plastics may creep.
- Loading rate: Did it break during impact or during slow bending? The failure mode may differ.
- Temperature: Is the object used near a polymer’s Tg (for PET, around 70°C, indicative)?
- Environment: Does it see humidity, UV, solvents, salt air, or heat cycling?
- Geometry: Are corners rounded or sharp? Are holes or cutouts creating stress concentration?
- Flaws: Are there scratches, chips, or sharp notches? Brittle materials are especially sensitive.
- Time under load: Is the part holding a constant load for long periods? Plastics may creep.
- Loading rate: Did it break during impact or during slow bending? The failure mode may differ.
- Temperature: Is the object used near a polymer’s Tg (for PET, around 70°C, indicative)?
- Environment: Does it see humidity, UV, solvents, salt air, or heat cycling?
- Geometry: Are corners rounded or sharp? Are holes or cutouts creating stress concentration?
Everyday failure checklist
- ✓Flaws: Are there scratches, chips, or sharp notches? Brittle materials are especially sensitive.
- ✓Time under load: Is the part holding a constant load for long periods? Plastics may creep.
- ✓Loading rate: Did it break during impact or during slow bending? The failure mode may differ.
- ✓Temperature: Is the object used near a polymer’s Tg (for PET, around 70°C, indicative)?
- ✓Environment: Does it see humidity, UV, solvents, salt air, or heat cycling?
- ✓Geometry: Are corners rounded or sharp? Are holes or cutouts creating stress concentration?
Multiple perspectives: why “make it stronger” isn’t always the answer
Consumers often demand “stronger materials,” while engineers often respond with “better design.” Both perspectives are valid, and both can fail.
- Increasing strength without improving toughness can produce a brittle part that still cracks from flaws.
- Increasing toughness may add cost, weight, or manufacturing complexity.
- Improving geometry and reducing stress concentration can yield outsized gains without changing material.
The honest conclusion is that durability is rarely a single knob. It’s an optimization among material properties, shape, environment, cost, and how people actually use the object—messily, repetitively, and sometimes carelessly.
- Increasing strength without improving toughness can produce a brittle part that still cracks from flaws.
- Increasing toughness may add cost, weight, or manufacturing complexity.
- Improving geometry and reducing stress concentration can yield outsized gains without changing material.
The honest conclusion is that durability is rarely a single knob. It’s an optimization among material properties, shape, environment, cost, and how people actually use the object—messily, repetitively, and sometimes carelessly.
Key Insight
Durability is rarely one knob. It’s an optimization among toughness, geometry, environment, cost, and real human use—messy, repetitive, and sometimes careless.
A more realistic way to look at everyday breakage
People want a simple story: the object was good or bad; the material was strong or weak. Materials offer a more interesting story. Objects fail because cracks grow, because plasticity either saves them or doesn’t, because polymers change with temperature, because time turns small loads into large deformations, and because environments quietly rewrite material behavior.
Fracture toughness values make the hierarchy plain: soda‑lime glass at roughly 0.7–0.8 MPa·m^1/2 lives in a different world than aluminum at roughly 14–28 MPa·m^1/2. Tg values explain why a PET-based plastic can become more vulnerable around 70°C, while polystyrene and acrylic sit closer to 100–105°C. Paris’ law gives language to the unsettling truth of fatigue: failure can be delayed, not avoided, until the final moment.
A good mental model is not “How strong is it?” but “How does it behave with flaws, over time, at this temperature, in this environment?” Ask those questions and everyday failure becomes less like betrayal and more like physics keeping its appointments.
Fracture toughness values make the hierarchy plain: soda‑lime glass at roughly 0.7–0.8 MPa·m^1/2 lives in a different world than aluminum at roughly 14–28 MPa·m^1/2. Tg values explain why a PET-based plastic can become more vulnerable around 70°C, while polystyrene and acrylic sit closer to 100–105°C. Paris’ law gives language to the unsettling truth of fatigue: failure can be delayed, not avoided, until the final moment.
A good mental model is not “How strong is it?” but “How does it behave with flaws, over time, at this temperature, in this environment?” Ask those questions and everyday failure becomes less like betrayal and more like physics keeping its appointments.
1) Why does glass shatter so easily if it’s “strong”?
Glass can be strong in compression, but everyday breakage is driven by cracks. With low fracture toughness—soda‑lime glass is often cited around 0.7–0.8 MPa·m^1/2—small surface flaws concentrate stress and cracks propagate with little energy dissipation. A tiny scratch can matter more than a large, evenly distributed load.
2) What’s the difference between strength and fracture toughness?
Strength describes how much stress a material can take in an idealized sense. Fracture toughness (K_Ic) describes how well a material resists crack growth. Real objects contain flaws, so toughness often governs real-world durability. That’s why aluminum (often cited 14–28 MPa·m^1/2) tolerates damage that would shatter glass.
3) Why can the same object fail differently depending on how it’s hit or loaded?
Materials can be rate-sensitive. Under dynamic loading, measured toughness can change; a Purdue study on soda‑lime glass reported values roughly 0.66–2.22 MPa·m^1/2 across a tested dynamic regime. Even when the material remains brittle, how quickly you apply load can alter whether cracks initiate, how they propagate, and what the failure looks like.
4) What is fatigue, and why does it make failure feel “sudden”?
Fatigue is crack growth under repeated loading. A component may survive many cycles, then fail abruptly once a crack reaches a critical size. Engineers often describe growth with Paris’ law: \( da/dN = C(\Delta K)^m \). The early stage is invisible, so the final break appears to happen without warning.
5) Why do plastics sag over time even when they aren’t overloaded?
Many polymers undergo creep—slow, time-dependent deformation under constant load. Temperature accelerates this. If the service temperature approaches the polymer’s glass transition temperature (Tg), chain mobility increases and creep can worsen. Indicative Tg values include PET ~70°C, polystyrene ~100°C, and PMMA ~105°C.
6) What role do humidity and UV play in everyday material failure?
Environment can amplify mechanical damage through environment-assisted mechanisms. Humidity, UV, solvents, salt, and heat cycles can weaken materials directly or degrade interfaces (adhesives, coatings), leading to interfacial debonding. The material may not be “weak”; it may be operating in conditions that steadily reduce its margins.
7) Are published toughness numbers reliable?
They are useful but not absolute. Toughness depends on test method and specimen details, which is why standards exist (engineers reference ASTM E1820 for J‑integral methods and ASTM E399 for plane‑strain K_Ic). For readers, the safest use is comparative: glass is far less tough than metals, even if exact values vary by test and formulation.
T
About the Author
TheMurrow Editorial is a writer for TheMurrow covering science.
Frequently Asked Questions
Why does glass shatter so easily if it’s “strong”?
Glass can be strong in compression, but everyday breakage is driven by cracks. With low fracture toughness—soda‑lime glass is often cited around 0.7–0.8 MPa·m^1/2—small surface flaws concentrate stress and cracks propagate with little energy dissipation. A tiny scratch can matter more than a large, evenly distributed load.
What’s the difference between strength and fracture toughness?
Strength describes how much stress a material can take in an idealized sense. Fracture toughness (K_Ic) describes how well a material resists crack growth. Real objects contain flaws, so toughness often governs real-world durability. That’s why aluminum (often cited 14–28 MPa·m^1/2) tolerates damage that would shatter glass.
Why can the same object fail differently depending on how it’s hit or loaded?
Materials can be rate-sensitive. Under dynamic loading, measured toughness can change; a Purdue study on soda‑lime glass reported values roughly 0.66–2.22 MPa·m^1/2 across a tested dynamic regime. Even when the material remains brittle, how quickly you apply load can alter whether cracks initiate, how they propagate, and what the failure looks like.
What is fatigue, and why does it make failure feel “sudden”?
Fatigue is crack growth under repeated loading. A component may survive many cycles, then fail abruptly once a crack reaches a critical size. Engineers often describe growth with Paris’ law: \( da/dN = C(\Delta K)^m \). The early stage is invisible, so the final break appears to happen without warning.
Why do plastics sag over time even when they aren’t overloaded?
Many polymers undergo creep—slow, time-dependent deformation under constant load. Temperature accelerates this. If the service temperature approaches the polymer’s glass transition temperature (Tg), chain mobility increases and creep can worsen. Indicative Tg values include PET ~70°C, polystyrene ~100°C, and PMMA ~105°C.
What role do humidity and UV play in everyday material failure?
Environment can amplify mechanical damage through environment-assisted mechanisms. Humidity, UV, solvents, salt, and heat cycles can weaken materials directly or degrade interfaces (adhesives, coatings), leading to interfacial debonding. The material may not be “weak”; it may be operating in conditions that steadily reduce its margins.
