Shot Blasting Concrete: When and How
Whether to blast is a different question from how, and it is already answered elsewhere. This is what happens once you have decided: the four things that have to be true before the machine can run, the three settings that decide the profile, and the one variable that changes under you while you work.
A shot blaster is set up rather than driven. The profile comes out of three settings you choose before you start: abrasive size, travel speed and shot flow. None of them predicts a result, so the profile is found on a test strip against a profile chip rather than calculated. And it does not hold still. The abrasive breaks down as the machine recycles it, so the same settings cut a shallower profile as the day goes on. Four things have to be true before any of that can happen: a dry floor, a clear floor, real power, and runs long enough to drive.
Whether to blast is a different question, and it is already answered
Grinding vs shot blasting settles the choice: what each method does to the surface, where each one honestly wins, why the answer on a commercial floor is usually both, and why the cost comparison is a crossover rather than a verdict. None of that is repeated here.
This page starts after that decision, because the operation itself is where blasting stops resembling grinding. A grinder is a machine you drive: you can see the cut, take another pass over a bad patch, and stop when the profile matches. A blaster is a machine you configure and then commit to. It applies whatever it was set to across everything it can reach, at the speed you walk it, and the surface it leaves is a consequence of decisions made before the first path was cut.
Four things that have to be true before it can run
Each of these can cancel the day on its own, which is why they belong in the site survey rather than in the first hour on site.
- A dry floor. Moisture clumps abrasive and fouls the recovery and separation the machine runs on. It is also where residual steel starts staining.
- A swept, clear slab. Loose debris goes straight into the recovery system. A separator sorts abrasive from dust; it is not a shovel.
- A real power supply. What that means is the specific machine’s requirement, not a general rule. Get it from the unit you are hiring and confirm it with the building.
- Continuous driveable runs. Not the floor area: the area the machine crosses without stopping and lifting. Racking, plant and doorways take large bites out of it.
The contamination case is worth naming separately, because it is the one that looks survivable and is not. Oil fouls abrasive recovery, and a blaster driven over a contaminated slab spreads the problem along its path in the same way a grinder does. Degreasing oil soaked concrete covers why that has to be finished, and the slab dried, before either machine starts.
The three levers that decide the profile
Depth and texture are outputs of these, and they interact, which is why changing two at once is how an afternoon disappears. Change one, run a strip, look at it.
| Lever | What it changes | When to reach for it | What it costs you |
|---|---|---|---|
| Abrasive size | The energy in each impact, so the size of every crater. | When the profile is the wrong texture rather than the wrong depth. | A charge change, so it is the slowest lever to undo. |
| Travel speed | How many impacts land on a given patch of floor. | First, for almost any depth problem. Slower cuts deeper. | Production time, and nothing else. Reversible on the next pass. |
| Shot flow | How much abrasive the wheel throws per unit of time. | To trim depth once the size is right and the speed is set. | Abrasive consumption, and it interacts with the other two. |
Reading across a row: the setting, what it physically alters, the situation that should make you reach for it, and what moving it costs. No sizes, speeds or flow rates are published here, because they are properties of a particular machine and a particular abrasive rather than of the method, and a number invented for a table would be worse than none.
Put a number on the speed you choose
Production here is geometry, not a judgement: the blast path width multiplied by the travel speed. Because the rate is derived rather than assumed, you can watch it move as you change the speed, which is how the profile decision gets a price attached to it.
Open floor only. Mobilization, turning, overlap and the edge grinding that follows are all outside this figure.
Open the live calculator →The test strip is the method, not a formality
There is no pass count to work from here, and that is the practical difference from grinding. A grinding schedule can be built from a rate and a number of passes before anyone arrives. A blast profile cannot be predicted from settings at all, because it depends on the abrasive that is in the machine that day and on the slab in front of it.
So the sequence is set, blast a strip, compare, adjust one thing, blast another strip. What you compare against is a profile chip, and the number on it comes off the coating manufacturer’s data sheet rather than off habit; CSP explained covers the scale and why over profiling is a failure in its own right.
Two disciplines make that loop worth anything. Take the strip somewhere representative, not on the cleanest, hardest patch you can find, because a soft area will blast deeper at the same settings and soft concrete is exactly where a blaster overshoots quietly. And leave the accepted strip unblasted where you can, so the reference stays on the floor instead of in somebody’s memory of it.
The profile drifts all day, and nothing looks different
This is the finding worth carrying off the page, because there is no warning attached to it. The machine sounds the same, the settings are untouched, and the floor is getting less profile than it was this morning.
A blaster recycles its abrasive continuously: it throws shot, recovers it with the debris, separates the two and throws the same shot again. Every one of those cycles breaks some of it and rounds off the rest, and the separator pulls the fines and broken pieces out as waste. What is left is a working mix that is steadily smaller and rounder than the charge you started with, and smaller, rounder shot cuts a shallower crater at the same speed and flow.
The consequence is about evidence. A strip approved at the start of the shift says nothing about the floor blasted after lunch, so a single verification at the beginning is not verification of the job. Add abrasive on a schedule rather than waiting for the cut to feel wrong, and re-run a strip partway through. On a floor that runs across a shift change, that mid point check is the difference between delivering one profile and delivering two.
Stripes you cannot see until the coating is down
A blaster cleans a path of a fixed width, and the profile at the edges of that path is not the same as the profile down the middle of it. How the paths are laid next to each other therefore ends up in the finished floor.
Too little overlap leaves a lightly blasted ribbon between passes. Too much lays a second full dose down a lane that already had one, cutting it deeper than the rest. Either produces a regular pattern running the length of the machine path, and the width of the pattern is the giveaway: a defect that repeats at exactly the machine’s working width is a tracking problem rather than a slab problem.
The timing is the cruel part. On bare gray concrete under work lighting, a correctly blasted lane and a slightly over blasted one look much the same. Gloss reveals texture, so the pattern arrives with the first coat that reflects anything, by which point the floor is coated and the fix is not a small one. Marking the paths out and holding the overlap consistent costs nothing on the day.
What is left to do once the machine stops
A blasted floor is not a prepared floor, and there are two items between them. Neither is optional and one of them is a separate line on the quote.
- Clear the residual abrasive. A fresh profile is open and sharp, so it holds stray shot and steel fines in the texture. Under a coating that is a weak point, and on a damp slab steel sitting in concrete can stain. Vacuum rather than sweep, and check the low points and joints where loose material collects
- Grind the edges to match. The blaster reached no wall, corner, column or anything it could not drive over. A grinder has to bring all of it to the same profile as the field, and an edge left different is both a visible line and a weaker bond
The edge work is slow relative to its area and it is the most commonly under quoted item in a prep package, so it is worth pricing as its own line rather than folding it into a field rate. The grinding time calculator takes an area, a rate for one pass and a pass count, and a perimeter strip runs at a very different rate from open floor. Then confirm the whole slab against the four prep gates, because a correct profile is only one of them and moisture is the one more machine time cannot fix.
Where to go next
If the method is still open, grinding vs shot blasting is the comparison, including why the honest version is usually blasting the field and grinding the edges rather than one or the other. For what the test strip is being measured against, the CSP reference chart lists the profiles and what typically produces them.
If the slab has anything on it, deal with that first: degreasing oil soaked concrete for contamination, and how to remove old epoxy and paint for a coating, which also covers why a soft film smears under abrasive instead of shattering. And if a finished floor has already failed and you are reading backwards, why epoxy floors peel separates an under prepared slab from the causes that look the same on the day.
Frequently Asked Questions
What actually decides the profile a shot blaster leaves?
Three settings, and none of them is how hard you push. A blaster is set up rather than driven, which is the main way it differs from a grinder.
The first is abrasive size. Larger shot carries more energy per particle and fractures a bigger crater, so it opens a coarser profile. Smaller shot spreads the same energy over more, smaller impacts and leaves a finer one. The second is travel speed. The slower the machine moves, the more impacts land on any given patch of floor, so the profile deepens. The third is the shot flow, meaning how much abrasive the wheel is throwing per unit of time.
They interact, which is why changing two at once is how people get lost. Change one, run a strip, look at it. The usual first move for a profile that is too shallow is to slow the machine down, because it is the one lever that costs nothing but time and is reversible on the next pass.
How do I know I have reached the profile the coating asks for?
By blasting a test strip and comparing it against a profile chip, because there is nothing to calculate. Unlike a grinding pass count, no combination of settings predicts a profile in advance: it depends on the abrasive in the machine that day and on the slab in front of you.
So the sequence is to set the machine, blast a strip somewhere representative, and hold the result against the concrete surface profile the data sheet specifies. Adjust one setting, blast another strip, and repeat until it matches. Representative matters here. A strip taken on the hardest, cleanest part of the floor tells you very little about a soft or contaminated area, and shot blasting is unforgiving on soft concrete because it keeps removing material at whatever rate it was set to.
Keep the accepted strip visible and unblasted if you can, so there is a physical reference on the floor rather than a memory of one.
Why is my profile getting shallower as the day goes on?
Because the abrasive is wearing out inside the machine, and this is the part that is easiest to miss because nothing about the machine looks different.
A blaster recycles its shot. Every pass through the wheel and off the floor breaks some of it down and rounds the rest off, and the separator pulls the fines and the broken pieces out as waste. What is left is a working mix that is gradually smaller and rounder than the charge you started with, and smaller, rounder shot cuts a shallower profile at the same settings.
So the profile drifts even though nothing was changed, which means a strip verified at the start of the shift is not evidence about the floor being blasted after lunch. The answer is to add abrasive on a schedule rather than waiting for the machine to feel wrong, and to re-run a test strip partway through the day. On a long pour that runs across a shift change, that check is the difference between one profile and two.
Why did my blasted floor come out striped once the coating went on?
Overlap. A blaster cleans a path of a fixed width, and the profile at the edge of that path is not the same as the profile in the middle of it, so how the paths are laid next to each other shows up in the finished floor.
Too little overlap leaves a lightly blasted or untouched ribbon between passes. Too much lays a second full dose of abrasive down a lane that already had one, cutting deeper there than everywhere else. Either way you get a regular pattern running the length of the machine path.
The cruel part is the timing. On bare gray concrete the difference between a correctly blasted lane and a slightly over blasted one is close to invisible, especially under work lighting. It becomes obvious once a coating goes down and reflects, because gloss reveals texture. Marking out the paths and keeping the overlap consistent is the fix, and it is far cheaper than the alternative, which is discovering the stripes in the first topcoat.
What has to be true before a shot blaster can run at all?
Four things, and each of them can cancel the day on its own, so they belong in the site survey rather than in the morning.
The floor has to be dry. Moisture makes abrasive clump and fouls the recovery and separation the machine depends on, and a damp slab is also where residual steel starts staining. It has to be swept and clear, because loose debris, grit and anything left on the slab goes straight into the recovery system, and a separator is not a shovel. The machine needs a real power supply, and what that means is the specific machine’s requirement rather than a general rule, so it is a number to get from the unit you are actually hiring and to confirm with the building before the day.
And the floor has to be driveable in continuous runs. That is the one people misjudge, because it is not the floor area. It is the area the machine can cross without stopping and lifting, and racking, plant, partitions and doorways take large bites out of it.
Is anything left to do after blasting, before the coating goes down?
Yes, and skipping it undoes the prep. Two things, and the second is a scheduling item rather than a task.
The first is clearing residual abrasive. A freshly blasted profile is open and sharp, and it holds stray shot and steel fines in the texture. Left there under a coating they are a weak point, and on a damp slab steel sitting in concrete can stain. Vacuum the floor properly rather than sweeping it, and pay attention to the low points and any joints, which is where anything loose ends up.
The second is the edges. A blaster cannot reach a wall, a corner, a column or anything it cannot drive over, so a grinder has to follow it and bring those areas to the same profile as the field. That work is slow relative to its area and it has to match, because an edge left at a different profile is both a visible line in the finished floor and a weaker bond. It is the most commonly under quoted item in a prep package.