Cracked concrete

BOF

we built a new house 3 years ago and had infloor heat installed. The system works well. This past year we built a 1500 sq foot hobby shop and had infloor heating installed. The day the plumber turned the heat on was very cold. He set the thermostats at 15 degrees Celsius. A few hours later I heard a loud boom and felt something in my feet. Looking around I saw a crack in my floor. Twenty minutes later there was another boom but not as loud. That evening I found multiple cracks in my floors. I have 2 levels in my shop and both floors have cracked. The concrete is over a year old and previously had no cracks. We have no cracks on any of the floors in our house. Does anyone have any insight on this or are there procedures for turning infloor heat on with cold concrete. Thank you

hot_rod

Has the supply water temperature to the slab changed from the previous years? What temperature are you sending to the slabs?

HVACNUT

How cold is cold?
0° F to 59° F is a shock.

GroundUp

This is why plumbers should not be doing hydronics. Any idea what the supply water temp is/was and the ambient temp of the slab that day?

BOF

I asked the plumber what the water temp was and he said 140F. The slab was probably below 40F.

Jamie Hall

No wonder the poor thing cracked. Now that it is you don't have to worry about it doing it again... which is the good news...

But the water should not have been anywhere near that hot -- and shouldn't ever be, for that matter.

Gordy

Concrete floors should be able to take 140 SWT, if properly acclimated. Depends on the radiant design requirements. Poor designs would require such a higher SWT.

The problem is the concrete was not brought up to temp slowly over a period of time.

It also depends on tube centers. Wide spacing creates huge DT areas across the slab.

You have to understand with 140 degree water the tubing expand in diameter, and length. The slab was to cold, and not strong enough to contain the expansion of the tubing.

Concrete has a coefficient of linear expansion/contraction @ 1/8” for every 100’ for every 10 degrees of temp change. Pex is far higher!

Gordy

A question I have is, where did the cracking take place?

Was there properly placed expansion / contraction joints in the concrete floor?

nibs

Sorry about the cracking, but hopefully it has not damaged the pex.
Lots of arguments in other threads here everyone seems to have an opinion.
No more than 120 deg for inslab radiant is the norm. For our slabs, while tiling, left the relevant loop off for 4 weeks, then cracked the ball valve just a bit each day for about 3 or 4 days, and that is on slabs that have been poured over a year ago and were heated last winter.
Made the mistake of heating my slabs too quickly when we first turned on the boiler, no cracking (FIBER, FIBER, FIBER) But we did get a little edge curl at the cast in control joints.
Few cement placers and finishers know about radiant slabs, few plumbers know about heating cement slabs, and lots of boiler installers don't know much about radiant slab concrete. Most engineers have no clue about radiant slabs.
Concrete expands and contracts with heat at about the same rate as steel, which is one reason steel is used as reinforcing.
Concrete has almost no torsional strength, so we use steel for the torsion strength, concrete excels in compressive strength, but if you heat a local area of a cement slab it is going to induce stresses as some parts want to expand and others do not.
Warm it slowly, Caution is much cheaper than repairs.

My experience with concrete began when I was 21 (55 years ago) working on a very large cast in place planetarium, in '76 worked on a ferrocement 65 ft yacht, have worked on more pours than I can count, have taught a course on some aspects of cement at a trades college. My company built a public building with electric infloor in 1995, built to engineers specs and the damn thing never worked well. Built ferrocement walls for my house, poured the thin slab radiant floor my house is all site mixed cement, using fiber, super plasticizer, fly ash, selected sand, coarse sharp sand for CIP and finer rounded river sand for ferro cement.
Don't take uncle joe's advice, 'cause he made the floor for his garage.
Cite engineering studies if you think my advice is in error.
The time to get it right is before it is built, then it goes easy.

Phew, sorry it was so long and strident, but concrete is raddled with people who know a little and give a lot of advice and it gets frustrating.

Gordy

My only point about concrete being able to take 140 SWT is that snowmelt does it all the time, yes?

An interior radiant concrete slab as you say should need less than 120 for a good design.

kcopp

How is the heat being regulated? How is it being heated? Do you have any pix?

BOF

I had the most experienced concrete guys do the pour. There was not a guy on the pour with less than 20 years experience. I paid a premium to get these guys. It was a 6 inch thick 32 Mpa mix with plastisers and wire mesh. There is 2 inches of insulation under the concrete with 6 feet of compacted gravel. The floor came out perfect. It is two engineered slabs with no saw cuts. That was the design. I spent $9000.00 polishing and sealing the concrete after it was 6 months old. The 2 cracked slabs are 500 sq ft and 950 sq ft. I did not have heat in the building last winter and there were no issues with frost or movement on the concrete.

The plumber was referred to me as having a lot of experience with in floor heat. He was told up front that the job was to be premium quality before he priced the job.

I have 4 zones each with its own thermostat. Two of the rooms have plank porcelain tile. There is no cracks on those floors.

Gordy

Please tell me what was engineered on a 6” slab that required zero contraction / expansion joints?

32 mpa is 4632 psi concrete

Plasticisers are there to increase slump with out reducing strength. They have zero to do with the final psi strength other than not having to add water to increase slump. Wire mesh does zero to prevent cracking. It just keeps the pieces together as a unit.

Did the designer know it was going to be radiant?

GroundUp

Sounds like this genius had better have some good insurance

kcopp

I would still like to know how the water is heated. Its possible the boiler malfunctioned and sent some super hot water into the slab. Was there a mix valve? Injection to the zone? Oil boiler? Water heater? Gas boiler?

hot_rod

Gordy said:

My only point about concrete being able to take 140 SWT is that snowmelt does it all the time, yes?



An interior radiant concrete slab as you say should need less than 120 for a good design.

And snowmelt doesn't generally come up slowly. Turns on at maybe 0° gets 100 plus SWT immediately. It could be at a rate of 200 BTU/ ft so 10 times the heat input a heating slab may see.

I think the key with SIM is the slab usually has somewhere to expand, it not restrained on all sides like a home basement slab often is. No place to expand, I've heard this called external restraint cracking.

nibs

@Gordy , Plasticizers are used in two ways, one if you add plasticizer and use the same amount of water, the cement flows much more readily is easier and cheaper to level, and cures to the same strength (psi).
Or you can use it to cut the water content so that the cement is stiffer harder to place and level, with less water the cement cures to a much higher strength, is less prone to shrinkage cracking and edge curl.
Google it and you will see that engineering studies have been done extensively.

Gordy

You are saying essentially the same thing.

Plasticisers increase workability by increasing slump, with out the addition of water to decrease strength. That’s it.

Plasticisers only last so long. If there is an extended haul time from far away Redimix plants, plasticisers should be added on site.

Also if loads sit to long on site before use.

Typically strict DOT, and ACI standards are 1 hour delivered and discharged from truck. 1 1/2 hours if retarder is added to the mix design.

Those standards are rarely adhered to in residential, and a lot of commercial jobs.


Water should never be added if the plasticiser loses its effectiveness. Only more plasticiser

Gordy

Here is one more note. Air entrainment. Air if tested (doubtful) should be between 5 and 8 % for all concrete that sees freeze potential.

Factory slabs are usually 0% since it is a controlled environment.

Zman

The only mistake I can see here was made by the person that designed it with no control joints and set the expectation that it would not crack.
@Gordy, good to hear from you. I was reading the beginning of this post and thinking "I haven't heard from Gordy in a while, he would like this post" and there you were!

Gordy

> @hot_rod said:
> (Quote)
> And snowmelt doesn't generally come up slowly. Turns on at maybe 0° gets 100 plus SWT immediately. It could be at a rate of 200 BTU/ ft so 10 times the heat input a heating slab may see.
>
> I think the key with SIM is the slab usually has somewhere to expand, it not restrained on all sides like a home basement slab often is. No place to expand, I've heard this called external restraint cracking.

Yes! When the foundation walls, and slab were poured all is acting as one unit under ambient temps everything moves at same expansion coefficient.

The minute you light off the slab the walls are lagging any warming the slab is seeing. Constricting it’s movement of expansion.

Now if there was an Edge insulation detail like we should see. Expansion has a chance.

Gordy

There are only 4 ways to “decrease” not eliminate the amount of control joints.

Post tensioning
Prestressed
Continuous reinforcement steel. “Not wire mesh”
Pins “ small tie wire size pieces of steel in mix” along the lines of fiber mesh, but much better.

If your roughly 30x30 and 23x23 areas had any wall projections into the slab not good.

realliveplumber

One of my best friends is a concrete guy.

2 things he will guarantee about concrete.

It will turn white.

It will crack.

nibs

@hot_rod as @Gordy points out air entrainment is gonna cut cracking in exposed environments such as snow melt.
(I forgot about that sorry)
When mixing our thin slab floor we added a few drops of detergent, which wets the water and helps long chain molecules and the foaming gives a little air entrainment for crack resistance.
But Fiber helps our digestive systems and really adds crack resistance to our concrete at a very low cost.

Gordy

> @realliveplumber said:
> One of my best friends is a concrete guy.
>
> 2 things he will guarantee about concrete.
>
> It will turn white.
>
> It will crack.

I always say, if you don’t want it to crack leave it in the Redimix truck.

Most people don’t realise beyond the visual obvious how much concrete cracks. “Micro cracking”

The only way to see them is wet the floor down with water. As it dries the last areas to dry are the micro cracks.

Gordy

From my perspective analysing the information given. I wouldn’t be so quick to throw the “plumber” under the bus.

In a 40 degree slab that 140 water cools pretty fast by the time it makes its way around a loop. Be interesting to know the RWT. How long it was running when it popped etc.

Zman

It would take a humungous boiler to bring a super cold slab up to temp too fast. Just because you set the boiler to 140, that doesn't mean it will get to temp today. Even snowmelt slabs with 125k btu/ft take hours to get up to temp.

nibs

Local internal temperature differences are going to cause stress and stress often leads to cracking.
An aside:
Have mentioned here that we sometimes used cement with ground up styrofoam in the mix. We made several blocks as an experiment. We use some of them as a sidewalk and after 5 or 6 years outside in weather ranging to 20 below, none have cracked, they do not wear well but could if the blocks were topped with ordinary cement 1/2" thick or less.

CBRob

Tekmar snow melt systems have a setting to prevent the shock to a slab.
http://www.tekmarcontrols.com/support/tekmarglossary/33-support/glossary/210-slab-protection.html

CBRob

> @Zman said:
> It would take a humungous boiler to bring a super cold slab up to temp too fast. Just because you set the boiler to 140, that doesn't mean it will get to temp today. Even snowmelt slabs with 125k btu/ft take hours to get up to temp.

With my new snow melt setup we see about a 22 degree Delta when the circs are at the lowest setting.
I have not seen any thing higher than 75 at the supply yet.
It might be the slab protection setting

hot_rod

Another interesting concept with SIM is the use of ∆P circulator in constant pressure setting. This allows the pump to ramp back as the glycol thins. So you don't end up over pumping the warmed glycol. Also some electrical energy savings.
eeping in mine places like Vail measure SIM in acres not square feet!

Mark Fryer out in Colorado showed that trick in our SIM webinar last week.

BOF

My house basement floor is in floor heat as well. It is 4000 square feet and we had no issues with it when it was turned on.
Thanks for all of your comments on this. I have turned it over to the insurance company to sort out.

Brian

Zman

Do you have pictures of the cracks? I think the expectation of no cracks is problematic. If the cracks are huge or displaced, that is a different matter.

pecmsg

FWIW
Floors in ice arenas are continuous pours with no expansion joints. So yes it can be done but with much more expense.

Jamie Hall

pecmsg said:

FWIW

Floors in ice arenas are continuous pours with no expansion joints. So yes it can be done but with much more expense.

And those are almost always fiber or pin reinforced as well as rebar -- and not that thin, either.

CBRob

Wouldn't it take 100s of BTUs per ft to heat concrete quickly?

With 120 BTUs per foot I'm seeing a Delta t of typically 20f

Hardly enough to shock a slab, right?

Gordy

Ice Rink floors are a completely different animal. There is perimeter expansion joints.

nibs

Fiber is so cheap and so effective, I would never pour without it.

Gordy

Fiber has its uses. Hard troweled decorative finishes is not one.

hot_rod

CBRob said:

Wouldn't it take 100s of BTUs per ft to heat concrete quickly?



With 120 BTUs per foot I'm seeing a Delta t of typically 20f



Hardly enough to shock a slab, right?

I question the a typical residential slab 20- 25 btu/sq ft. could be heated fast enough to cause cracking. The sun shining in a large window wall might warm a slab as fast as the radiant.

Certainly no harm in ramping it slowly if you can babysit the job.

So many variable involved with concrete cracking
sub grade compacted properly? Consistent slab thickness? Poured without additional water added? Room for expansion on all sides?
Adequate thickness over the tube, 3 times the aggregate size. Stop and start pour? Admixed used? Properly? Was it fogged or sealed? Too much power troweling (to burn off fibers sticking through

A crack is a free control joint.

rick in Alaska

I for one would like to see a picture of the crack along with a layout of the tubing. Seems to me if the tubing was running too hot and caused the cracking, which I doubt, that it would follow the tube.
Rick