Concrete Strength and the Water-Cement Ratio

Every extra pound of water makes concrete easier to place and weaker once it hardens.

A ready-mix truck pulls up to a pour, and the concrete is too stiff to place. The driver reaches for the hose and adds several gallons per cubic yard. Five minutes later the mix flows beautifully. It looks better than when it arrived.

Nothing on site tells you what that water cost. The answer comes 28 days later, when a lab crushes a cylinder cast from that load. This Lesson shows why the news is bad, how the crushing test works, and how to put a number on the hose.

Rock, glue, and leftover water

Concrete is rock held together by glue. Sand and gravel, together called aggregate, fill most of the volume. The glue is cement paste: portland cement mixed with water. In ordinary concrete the rock is the stronger part, so the concrete is only as good as its paste.

Paste doesn't harden by drying out. Cement reacts with water, a reaction called hydration, and each grain grows solid crystals outward until they knit into their neighbors. That's why concrete can harden underwater. So water is a real ingredient, but cement needs surprisingly little of it. Full hydration takes about 0.4 lb of water per pound of cement, and a mix that lean is stiff and hard to place. So we add more water than the chemistry asks for, just to make the mix flow.

That extra water has nothing to react with. Once the concrete hardens, it leaves pores behind: holes where nothing carries load.

W/C = 0.35 W/C = 0.70 Fresh Hardened cementgrain water hydrationproduct pore
The same cement in two pastes
Schematic, not to scale. More water spreads the same grains apart. Hydration makes about the same amount of solid either way, so in the wetter paste it can't fill the gaps, and the leftover space becomes pores.

The picture shows why the number that matters is a ratio. How far apart the grains start depends on how much water each pound of cement gets. Double the water and the cement together, and the grains sit just as far apart: you've made more of the same paste, not a weaker one. Add water alone, and they spread out. So we track the water-cement ratio:

W/C=weight of waterweight of cementW/C = \frac{\text{weight of water}}{\text{weight of cement}}

Mix 300 lb of water with 600 lb of cement and W/C = 0.50. It's a ratio of weights, not volumes, so water measured in gallons gets converted first, at about 8.34 lb per gallon. The Handbook's Concrete section, in its Materials Science chapter, calls W/C "the primary factor affecting the strength of concrete." It states that effect without the reason, so the pores are yours to remember.

What the cylinder test measures

Nobody can see pores on a job site, so we measure what they do: crush some concrete and see what it takes. Before trusting the number, ask what you'd ask of any materials test: what does the test measure? The same concrete gives different numbers at different ages, under different curing, in different shapes. A strength means nothing without its test conditions.

The specified compressive strength, fc′f'_c, is the strength the designer writes on the drawings, in psi, and it's defined by one standard test:

  • Specimen: a cylinder, either 6 × 12 in. or 4 × 8 in. (diameter × length).
  • Curing: kept moist until the test.
  • Loading: axial compression until it breaks.
  • Age: 28 days.
  • Result: a strength test is the average of two or three companion cylinders from the same sample, not any one cylinder.

There's nothing magic about 28 days, since concrete keeps gaining strength after that. It's an agreed date, so results from different labs and jobs can be compared. The cylinder isn't the structure either. It's a yardstick. The Handbook's reinforced concrete section defines fc′f'_c only as "compressive strength of concrete (psi)" and doesn't print this procedure, so these conditions are also yours to remember.

Computing the strength is the easy part. It's the stress at failure, load over area, which is the Handbook's σ=P/A\sigma = P/A from Mechanics of Materials. The only real decision is which area, and the load comes down through the cylinder's circular end:

f=PA=Pπd2/4f = \frac{P}{A} = \frac{P}{\pi d^2 / 4}
P P d L loaded area, A = πd2/4 Standard sizes (d × L): 6 × 12 in. or 4 × 8 in.
A cylinder in axial compression
The load crosses the circular end, so that circle is the area that resists it.

It isn't diameter times length. That rectangle is what the cylinder looks like from the side, and the load doesn't push through it. Watch the units too: pounds over square inches gives psi, while kips give ksi.

Worked Example A strength test from two cylinders
1 of 3

Two companion 6 × 12 in. cylinders, moist-cured and tested at 28 days, fail at 118 kips and 124 kips. Find the result of the strength test, in psi.

Now the conditions behind those numbers. Only one of these reports is a fair comparison.

Checkpoint

Five lab reports come in for the same pour. Which one gives a result you can compare directly with fc′f'_c?

A.
One 6 × 12 in. cylinder, moist-cured, loaded axially at 28 days
B.
Two 6 × 12 in. cylinders, left in the open air beside the pour, loaded axially at 28 days, results averaged
C.
Three 4 × 8 in. cylinders, moist-cured, loaded axially at 28 days, results averaged
D.
Two 6 × 12 in. cylinders, moist-cured, loaded axially at 7 days, results averaged
E.
A 6 × 6 × 20 in. beam, moist-cured, broken in bending at 28 days

Two rules for accepting concrete

Well-made concrete still scatters. Two strength tests from the same mix won't match, so demanding that every test beat fc′f'_c would reject good concrete. What you really want to catch are two different failures: concrete that's running low overall, and one load that's badly off. Acceptance uses one rule for each:

  1. Every average of three consecutive strength tests is at least fc′f'_c.
  2. No strength test falls below fc′−500 psif'_c - 500\ \text{psi}, a floor under every single test.

These come from the ACI building code. The Handbook doesn't print them, so a question that uses them states them. They apply to strength tests, never to single cylinders. And "three consecutive" is a window that slides along the series: tests 1 to 3, then 2 to 4, then 3 to 5. One average of the whole series can hide a low stretch.

3,000 3,500 4,000 4,500 5,000 1 2 3 4 5 6 7 8 9 Strength test number Strength (psi) f'c = 4,000 psi f'c − 500 = 3,500 psi Test 3 is below the floor Tests 6–8 average 3,900 psi single test average of latest three tests
Both rules on one series of strength tests, f′c = 4,000 psi
Test 3 falls below the 3,500 psi floor, yet its neighbors keep every average that includes it above 4,000 psi. Tests 6, 7, and 8 all clear the floor, yet together they average only 3,900 psi. Each rule catches what the other misses.
Worked Example Checking a series of six tests
1 of 3

A job has fc′=4,500 psif'_c = 4{,}500\ \text{psi}. Its six strength tests, in order, are 4,900, 4,450, 4,400, 4,550, 4,800, and 4,650 psi. Under the two rules above, is the concrete acceptable?

From ratio to strength: the Handbook chart

The Handbook's Concrete section plots what all those pores add up to: average 28-day compressive strength against W/C.

NO ADDED AIR RECOMMENDED PERCENT ENTRAINED AIR 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 W/C BY WEIGHT AVERAGE 28-DAY COMPRESSIVE STRENGTH, PSI
Average 28-day compressive strength against W/C
The upper line is concrete with no added air. The lower line is concrete with the recommended percent of entrained air.

Source: Adapted from the FE Reference Handbook, Materials Science/Structure of Matter, Concrete, which credits the Concrete Manual, 8th ed., U.S. Bureau of Reclamation, 1975. Gridlines and labels added.

Strength falls steadily as the ratio climbs, with no kink at 0.4. Below that, some cement never reacts, but an unreacted grain is still solid, so less water still means fewer pores. The strength axis is logarithmic, which is why its gridlines crowd together toward the top, and why both lines plot straight. Entrained air means tiny bubbles added on purpose for durability. At the same W/C it gives less strength, so check which line a question is asking about. Workability, Air Entrainment, and Admixtures explains why the air costs strength.

This copy has extra gridlines and labels. The Handbook's own labels only 1,000, 2,000, 4,000, 6,000, and 8,000 psi, with W/C gridlines every 0.20, so expect to estimate on the exam.

From a target strength, go across to the right line, then down to W/C. A question then usually gives the mixing water, and the definition of the ratio gives the cement:

cement=waterW/C\text{cement} = \frac{\text{water}}{W/C}

With W/C below 1, cement always outweighs the water. If your cement comes out lighter than your water, you've multiplied instead of divided.

Worked Example Cement for a target strength
1 of 2

A mix with entrained air needs an average 28-day strength of 4,000 psi, and it has 270 lb of mixing water per cubic yard. Find the cement content.

NO ADDED AIR RECOMMENDED PERCENT ENTRAINED AIR 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 W/C BY WEIGHT AVERAGE 28-DAY COMPRESSIVE STRENGTH, PSI
Average 28-day compressive strength against W/C

Source: Adapted from the FE Reference Handbook, Materials Science/Structure of Matter, Concrete

The chart also runs the other way, from a batch to an expected strength. Count every pound of water that ends up in the paste. That includes any free water a question says the aggregates carry: the film on the outside of the stone and sand joins the paste as surely as water from the hose. Moisture Corrections to Batch Weights shows how to compute it. Divide the total by the cement, go up to the right line, then across to strength.

Worked Example Expected strength from a batch
1 of 3

A cubic yard of concrete with no added air is batched with 500 lb of cement and 230 lb of water. The aggregates carry another 45 lb of free water. Find the expected 28-day strength.

NO ADDED AIR RECOMMENDED PERCENT ENTRAINED AIR 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 W/C BY WEIGHT AVERAGE 28-DAY COMPRESSIVE STRENGTH, PSI
Average 28-day compressive strength against W/C

Source: Adapted from the FE Reference Handbook, Materials Science/Structure of Matter, Concrete

One ratio, four consequences

Most of what W/C does comes back to those pores. More water per pound of cement means more pore space, and that does three things:

  • Strength drops, since there's less solid to carry load.
  • Permeability rises, since the pores link into channels that water can travel through.
  • Durability suffers, since water and what it carries, such as deicing salts, can get inside.

The fourth runs the other way. Water is what makes the mix flow, so more of it means better workability: easier to place and finish. That's the whole temptation of the hose. Slump, the field measure of how stiff fresh concrete is, rises when you add water, and it tells you nothing about what just happened to W/C. The Handbook puts it this way: "Water content affects workability. However, an increase in water without a corresponding increase in cement reduces the concrete strength." Workability, Air Entrainment, and Admixtures covers slump, and how to buy workability without paying in strength.

Worked Example What the hose costs
1 of 4

A mix with no added air is batched with 600 lb of cement and 300 lb of water per cubic yard. On site, the driver adds about 7 gallons per cubic yard (60 lb) to restore slump. Find W/C before and after, and the expected 28-day strength each way.

NO ADDED AIR RECOMMENDED PERCENT ENTRAINED AIR 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 W/C BY WEIGHT AVERAGE 28-DAY COMPRESSIVE STRENGTH, PSI
Average 28-day compressive strength against W/C

Source: Adapted from the FE Reference Handbook, Materials Science/Structure of Matter, Concrete

Recap

  • Cement needs far less water to hydrate than it takes to make concrete flow. The excess leaves pores, so the ratio W/C, by weight, sets strength rather than the amount of water or cement alone.
  • fc′f'_c comes from moist-cured 6 × 12 in. or 4 × 8 in. cylinders loaded axially at 28 days, with a strength test as the average of companion cylinders. Each cylinder's strength is P/(πd2/4)P/(\pi d^2/4).
  • Acceptance works on strength tests: every average of three consecutive tests at least fc′f'_c, and no test below fc′−500 psif'_c - 500\ \text{psi}.
  • On the Handbook chart, pick the right line first. Cement is water divided by W/C, and W/C counts any free water the aggregates carry.
  • Raising W/C improves workability but lowers strength and durability and raises permeability. Adding water on site to restore slump pays that price.