WHY IS MY
NEW CONCRETE
CRACKING?
A crack reports that movement happened. It does not report why, and the difference between a slab that moved and a slab that is failing lives in the mix that was specified, the conditions on the day it was placed, and how it was cured, none of which are still visible in finished concrete. That is not a reason to panic and it is not a reason to look away. It is a reason to go after the record while the record still exists.
A crack is a symptom. It is not the event, and treating it as the event is what sends most homeowners down the wrong path.
The same visible crack can sit on a slab that will serve for decades and on one that is already failing. What separates those two slabs is not the crack. It is what went into the concrete, what the weather did to it while it was still workable, and what was done to it in the days after the crew drove away. Those facts were live for a few hours and then they stopped existing, so either somebody wrote them down at the time or nobody can produce them now.
A crack carries real information. It just does not carry the one you came for.
Start with what you can legitimately read off a slab, because it is more than nothing and a lot of advice online treats it as either everything or worthless.
Timing carries information. Something visible within hours of finishing describes a different event from something that shows up in month four, which in turn describes a different event from something that opens in year three. When you first saw it is a fact worth writing down before you forget it.
Location carries information. Inside corners, the points where a slab changes width, and the edges of anything set into the concrete all concentrate movement. A line that runs corner to corner across a bay is describing something different from one that follows a joint.
Whether it follows a joint carries information. Slabs are jointed because concrete changes volume as it cures and as it dries, and a joint is a decision about where that movement is allowed to go. Movement that arrived where somebody planned for it to arrive is a different event from movement that went somewhere nobody chose.
Change over time carries the most information of all. One photograph is a data point. The same view photographed again in 90 days, from the same spot, with something in frame for scale, is a measurement. It is also the single most useful thing you can start today, and it costs nothing.
What none of that reaches is the mix. Timing, location and movement describe the behaviour of the concrete. Whether that behaviour is ordinary or the leading edge of failure depends on what the concrete was made of and what was done to it, and those are questions about a day that has already passed.
Cracking is one of three things. The other two are usually worse.
Most people arrive at this question with one word for what is happening to their concrete. There are three in common use, they describe different mechanisms, and the one everybody knows the name of is not the one that ends slabs.
The one you searched for, and the one with the widest range of meanings behind it. It runs from movement that a joint was cut to receive, all the way to a genuine structural problem, and the visible line is a poor guide to which end of that range you are on. Defined once, with its sources, at the PSIP dictionary.
The surface breaking down and coming away rather than splitting. It is a different mechanism from a crack and most homeowners do not have a name for it until somebody supplies one, which is part of why it goes unaddressed longer. Which symptom a slab shows is driven by exposure, meaning the environmental conditions the concrete has to survive, and Sun Belt exposure is not northern exposure. ACI 318 and ACI 332 define those categories. Defined at the PSIP dictionary.
The most serious of the three, and the reason the other two are worth taking seriously rather than waiting out. It is an internal mechanism rather than a surface one, which means by the time it is legible on the outside it has been running for a while. It has its own page, /concrete-cancer, and its entry at the PSIP dictionary.
Four things decided this slab. All four stopped being visible when it set.
Concrete arrives as a liquid, is finished by hand, and passes out of the crew's hands within hours. Much of what governs how it performs over its service life is decided inside that window, and it leaves no trace the eye can read.
What was specified
The exposure class the mix was designed against. The governing codes are ACI CODE-318-25 for structural concrete and ACI CODE-332-20 for residential concrete. A finished slab does not show which class was chosen, or whether one was chosen at all.
What actually arrived
The batch ticket is the ready mix producer's record of what was in the truck, and it is the one third party document that already exists on an ordinary residential pour. Whether it is still retrievable depends on whether anyone kept it, which is the part worth acting on today.
The conditions on the day
Air temperature, concrete temperature, humidity, wind, how long the truck waited and when finishing started. These are live facts for a few hours and then they stop existing. The weather half of them is a public record and is the one variable you can still reconstruct.
How it was cured
What was done to the slab after the finishers left, and for how long. Curing is often overlooked or cut short, particularly when budgets and timelines are under pressure, and it happens after the crew that impressed you has gone.
That is why concrete documentation works differently from documentation in most trades. In most work you can go back and look. With concrete, going back and looking shows you the outcome and says very little about the cause. The moment documentation has to happen is the point of pour, and there is no later.
Six things you can recover this week, most of them free.
If the pour was recent, more of the record survives than people assume. Work down this list in order, because each one gets harder to obtain the longer you wait, and the first two are the ones that disappear.
Where it lives: the invoice, the permit file, your bank record, the text thread where somebody said they were coming Tuesday, or the timestamp on a photo you took. Why first: every other item on this list is retrieved by date. Without it you are asking people to search their records for "sometime in spring."
Ask plainly and without accusation, because this is a routine document and treating the request as an accusation is what makes it go badly. Say you are putting your home records together and you would like the ticket from the pour. Why it matters: it is the producer's record of what was actually delivered, made by a third party, and it either exists or it does not. The definition is here.
Why it can work: the plant that batched the load may hold its own copy, and a plant's retention practice is its own rather than the contractor's. What you need in hand first: the date and the delivery address. This is the main reason item one comes first.
Where it lives: the National Centers for Environmental Information publish archived daily observations by station, and it is a public record you can retrieve for free. What it gives you: the one pour day variable that was never in anyone's private custody. Temperature and wind on the day of placement, and what happened over the days the slab was curing.
Why it is not always obvious: on a job that ran through a builder, a broker or a subcontracted crew, the company you paid may not be the party who made the calls on pour day. What you want is a name, because the contractor of record is the party who owned those decisions and therefore owns the record of them.
What to do: photograph what you are seeing from a fixed spot, with a coin or a tape in frame for scale and the date visible or noted. Then do it again in 90 days from the same spot. Why it is worth more than it looks: whether something is stable or moving is the question every professional will ask you, and it is a question only a dated series can answer. If this ever becomes a dispute, you will be the party who has one.
Nothing on that list requires you to have known any of this before the pour. It is the recovery version of a record that should have been assembled on the day, and the reason PSIP exists is that the recovery version is harder, slower and less complete than the original would have been.
We score the record. We do not grade the concrete.
This page would be worth less if it ended by telling you the answer is us. So here is exactly what PSIP is and what it is not, in the same terms as everything above.
- ×PSIP cannot tell you why your slab did this. Nobody can answer that from a photograph, and a page that offered to would be selling you something. What a complete record does is turn the question from a guess into a matter of fact, which is a different service and an honest one.
- ×PSIP does not say a certified contractor does better work. A PSIP certified contractor has committed to documenting the pour and to retaining the record. That is a statement about their paperwork, not a judgment of their concrete, and treating it as the second thing would make it worth nothing.
- ×PSIP does not rank contractors. Records are sorted, never ranked. Sorting is an order you choose and can reverse. Ranking is a judgment made on your behalf, and a body that scores documentation has no basis for making it.
- ×A record is not a warranty. Certification is a documentation and scoring credential. It is not a warranty, guarantee, or insurance of the contractor's work; not engineering or design services; and not a government license or code approval. PSSG does not supervise, direct, inspect, or perform the contractor's concrete work, and is not a party to any contract between the contractor and the customer. The contractor is solely responsible for its own work, methods, safety, code compliance, licensing, and customer relationships. A closeout packet establishes what was specified and what was delivered, which is the factual ground a warranty or a claim would be argued from.
- ×Concrete still fails for reasons nobody controlled. Ground movement, loading, drainage and time are all real. Documentation records the variables that were controllable on the day. It does not remove the ones that were not.
If you are reading this because of something you found on your own driveway, the useful takeaway is section IV, and it works whether or not PSIP is ever involved. If you are reading it before hiring anyone, the useful takeaway is that everything section IV asks you to reconstruct is available for free at the time of the pour, from a contractor who is in the habit of writing it down.
Asked and answered.
A crack reports that movement happened. It does not report why, and the difference between a slab that moved and a slab that is failing lives in the mix that was specified, what actually arrived on the truck, the conditions on the day it was placed, and how it was cured. Those variables stopped being visible the moment the concrete set, which is why the answer is found in the record from pour day rather than in the crack.
Concrete changes volume as it cures and as it dries, and slabs are jointed because that movement is anticipated and given somewhere to go. So a line on a new slab is common, and on its own it is not evidence that the concrete is failing. What settles that question is the mix, the conditions on the day and the cure, none of which can be read off the surface.
Timing is information rather than a verdict. Something visible within hours of finishing describes a different event from something that appears in month four, which describes a different event again from something that opens in year three. The more useful move than deciding today is to record what you are seeing, from a fixed spot with something in frame for scale, and photograph the same view again in 90 days. Whether it is stable or moving is the question a professional will ask, and only a dated series answers it.
Not on its own. The same visible crack can sit on a slab that serves for decades and on one that is already failing, and what separates them is what went into the concrete and what was done to it rather than the appearance of the line. Cracking is one symptom of failure. Spalling and scaling is another, and alkali silica reaction is the most serious of the three.
Not reliably, and the honest ones say so. An experienced eye can narrow the possibilities from the timing, the location and whether the movement followed a joint, which is real value. What visual inspection does not reach is the mix that was specified, what was actually delivered, the conditions during placement and how the slab was cured, because those left no readable trace in the finished concrete.
They are three different mechanisms and PSIP treats all three as symptoms of the same event, which is concrete that fails. Cracking is splitting. Spalling and scaling is the surface breaking down and coming away. Alkali silica reaction is internal rather than surface level, which is why it is legible from the outside only after it has been running for some time. Each has a standalone definition in the PSIP dictionary, and alkali silica reaction has its own page at /concrete-cancer.
Often, and it is worth asking. The batch ticket is the ready mix producer's record of what was actually in the truck, so there are two places it can live: with the contractor, and with the plant that batched the load. A plant may hold its own copy on whatever retention schedule that plant uses. What you need before you call either one is the exact pour date and the delivery address.
Pin down the exact pour date, because everything else is retrieved by date. Ask your contractor for the batch ticket, plainly and without accusation, since it is a routine document. Pull the archived weather for the pour date and the days after, which is a free public record. Establish who the contractor of record was. Then photograph what you are seeing from a fixed spot with something in frame for scale, and repeat it in 90 days.
Every term on this page is defined once.
Failure, cracking, spalling and scaling, alkali silica reaction, exposure class, batch ticket, closeout packet, contractor of record and point of pour each have a single standalone definition in the PSIP dictionary, with the primary sources behind them.