Newb want to detect crack development in concrete.

Projects General Guidance
1

Hello Arduino experts,

I need a help regarding my project of detecting a crack in concrete.

The project has two parallel concrete beams each 500 ft long, 18 inch wide, and 10.5 inch deep. As this is pretty long structure and exposed to the environment, it will have some hairline cracks in random spots. My aim is to detect where the crack occurred first and when did it occur in a small segment (25 ft) of the beam.

I am a civil engineer (but I have moderate skills in programming and biginner-moderate skills about electronics). I would appreciate if anyone could please help me find sensor or way to detect these hairline cracks (in the order of microns).

My initial ideas are:
Idea 1:
Using an arduino with multiple parallel connection. each of these parallel wires would then be attached to the side of the concrete beam (as shown in attached figure). Each arduino would measure the crack development in 25 ft section (with five 5 ft wire placed as seen in figure). In ideal case, after the beam has hairline crack, the wire would break and the circuit would open. From the location of that wire, I would know where crack occured and the arduino would have to record the time when that circuit opens.

One problem with this method might be,the wire might not break at all with that small crack.

Idea 2:
Using some acoustic sensor or infrared sensor or laser and measure the distance change. Sensor would be attached to the side of concrete with reflecting device/surface at some regular distance (as shown in attached figure). The sensor should then record the time when the crack occured by measuring the movement. This idea might have a lot of disturbances picked up as the beam is outside. In addition, the crack is so small (in order of micron) that the resolution of these sensor might not be good enough.

This is the first issue. The other things that arduino must do is record that time in SD memory card and transfer it through wifi. However, those would be just other standard shields being used.

I hope that I have made my problem clear. Please let me know if any more infromation is required. I really would appreciate any help.

Thanks,
S

IMG_20150828_121349.jpg
IMG_20150828_121349.jpg2048×1536 363 KB

IMG_20150828_121410.jpg
IMG_20150828_121410.jpg2048×1536 420 KB

2

warning:
That are serious beams of concrete, I expect for building bridges or houses.
Do not expect to create a system that you can officially label the beams as crack free.
There are professional certified tools in the market for such analysis.

That said: cracks detected are cracks detected.

you need to investigate transducers and measure sound conductivity/reflection or so. Then have a system to learn the difference between OK and cracks. not trivial imho.

seen similar questions before once or twice but never final solutions

3

Rather than wires which might stretch sufficiently to span a tiny crack what about lines of conductive paint ?

But, in either case, how could you be sure you have the wires or paint in the right place ?

Could the cracks be detected with a camera and image processing software ? - far beyond the capacity of an Arduino.

...R

4

first off, if you have a wire that is taught, and stretch it to have it snap, the normal heating and cooling of the wire will be far in excess of the distance moved on a crack.

if you have a sonic ping, the crack will not carry the wave. you can use a robot to run the length and ping for cracks. this would be much more active crack detection.

you did not mention if the cracks are only a few mm deep or if they are much deeper. shallow cracks would need sensor and receiver close, deep cracks would allow for more distance.

5

I forgot to mention the use of these things. So, basically these beams are in open space. 500 ft long is pretty long beam— these are casted to represent the concrete pavement. So, we are only interested in surface crack as it is related to concrete road. While driving, people only care about the crack on road, as long as the crack on surface is tolerable, it is alright. We also have a section of 1.6 km of road in Chicago with vehicles moving at really high speed that we are interested to detect some cracking of concrete in — but before that, we want to test it in sample beams.

So, this project just needs to measure the crack seen on the surface. In fact, it does not need to measure the crack length or width, it just need to find the time when the crack occurs for the first time on the surface. Assuming that the crack on surface of road would also propagate through the side, it would be easier to detect the crack from side of the concrete as instruments would not be affected by traffic (or in other way, instruments would not disturb traffic).

We have experience of concrete pavement from long time, so although we don’t know where the cracks will occur, we know some regular pattern of transverse crack on the concrete road occur at around every 10 feet. So, we don’t need the exact location where the crack occurs, we just need approximate time when it occurs.

Another thing is, if we take a section of 25 feet and put 5 wires each five feet long on side of the concrete road, we will at least get one crack (or perhaps two cracks).

To ensure that the wire breaks, we can do one thing. We can glue a very thin high gage wire on side of the concrete as an integral part of concrete. If the concrete and wire are sticked together on side, when the concrete tries to crack, the wire would either deform or break. This would, either make the circuit open or change voltage (or current - I’m not exactly sure which one would be easier to measure with arduino) in that parallel section. All we want is to record the time when one of the parallel circuits is opened or experience those fluctuation (we can keep some threshold value change detection). I would appreciate if someone could suggest the way to make such system.

@Robin2 — We have done camera and image processing in the lab and it works pretty good. However, we are planning to implement this project on real highway. So, first thing is leaving camera on highway is not an option— as people might just steal the camera and we will lose the data. Also, powering them would be a problem in the real site. The third problem is, if the crack occurs at night, we would need to have proper lighting system — and lighting 25 feet of road edge in middle of highway and using the camera to record it would just not be possible — too much work for some people at night + also the same problem of power supply at middle of highway.

So, basically, the system I want would have a bunch( to be exact—>5) of parallel lines with very thin wire section of 5 feet in each branch. When either one (or more) of these wires break (or resistance change due to deformation), arduino should record the time (and store in SD card and send over wifi). From the location of which wire broke or deformed, we can know where (which 5 feet section of our road) the crack formed. Also, we can go and see the crack next day, it’s not a big deal about the location — all we are interested is when the crack forms.

I hope that the problem is more clear now. Sorry for any confusion.

I would appreciate if anyone can help me in providing some guidance on if it is feasible or how it can be made. Also, any information about video/ forum/ write up about how such circuit be connected would be great.

Thanks in advance.

S.

6

Hi.
Have you thought of adhesive strain guages?
If you use wire, it will have to be very brittle wire, hard drawn, so it doesn't stretch before breaking.

Tom.... :slight_smile:

7

Your best bet is to develop a very brittle paint on conductive material that can be painted on in very thin strips along the area that is to be studied. The idea being that if the concrete develops a crack, the conductive material will also develop a crack, which could be detected in a change in its resistance or other electrical property.

8

You forgot to mention it must also be impervious to salt water! Your concrete would not have the problem if the DOT quit using salt in the winter.

Paul

9

I am a geophysicist, and i would recommenced that you use an official, company made GPR to inspect the beams for hairline fractures... The amount of coding, data processing, and the speed in which all of this needs to be done is probably out of reach of the arduino

10

Concrete science is a world into itself. Where I work when concrete pours are made, sample cylinders of each pour are made, cured, then sent to a lab for testing to verify the concrete has the properties specified in the design documents. You may want to research existing concrete testing methods. As to detecting when a crack develops under test conditions, you might apply the stress slowly and use some type of surface optical monitoring. Mounting and then loading a large beam until it starts to develop cracks will be a big project itself.

11

@Tom

Commercially available strain gages would not work for this purpose. We are using those to embed inside the concrete to measure strains inside the concrete.
Brittle wire is good option.

@zoomkat

I liked the idea of brittle paint. Do you have any suggestion of such paints? I would love to try those in lab first and implement in the field.
Are there any examples of such conductive paint used on concrete? I would appreciate if you could provide me some references.

@Paul

I don't think that would be an issue in our case because-

  1. Concretes crack in the matter of few days. So, we can choose summer time to cast the concrete (or anytime before winter) and run this arduino for a couple of days.
  2. Even if it is winter. We would have control of those section of road. We can stop putting salt on that particular section of the Tollway.

@nrapp

GPR would be NDT test that we will do in aged concrete. Right now, we are only interested on surface crack on early age cement concrete pavement.

General questions:

  1. What is the maximum gauge that can be used with arduino? Can I use like 34 gauge wire (which I will stick firmly on the side of the concrete ) and measure the electric property to find the crack -- assuming that thin wires would have higher probability of breaking?

  2. Can fiber optic wires be used in some way?

  3. Can thin copper (or any other metal) foil be used to conduct on the section we are interested in? We can stick the foil on side of the concrete (and hope that it cracks as the concrete cracks)?

  4. Can conductive paints be used on concrete?

Thanks in advance. :slight_smile:

12

If you're looking for a conductor you might try carbon fiber (tow). Reliable conductivity and it certainly doesn't stretch.

13

@Chagrin

Looks like it is reliable with respect of conductivity. However, it seems like it will not break easily.

"This tow (or yarn) is composed of 24,000 individual carbon filaments, which boast the highest ultimate tensile strength in the industry. "
-Source

Can some conductive epoxy be used on side of concrete to make a circuit? This, probably, might be the best shot as the epoxy (or even paint) would crack with concrete.

14

You apparently found a "24K" tow, which is 24,000 fibers. I know you can find it down to 1K but even that might be a bit much.

Epoxy is pretty stretchy, or at least the types I've used.

You've certainly got a tricky problem to solve.

15

Thanks everyone for all the comments. It helped me.

I have not exactly solved this problem yet. But, I'm doing some experiments with a copper foil tape and a 30 AWG wire. I feel like I'm on right path.

I attached them to a DCT samples made of concrete (if you want to know about the sample - document). Copper foil tape was too wide for the crack to propagate. So, I had to cut the width to 1/5th. Still the failure was in adhesion. Then I tried a few samples of those cut foil tape with superglue on top -- IT WORKED :).

I also tried putting tension on the 30 AWG wire and glued it to the concrete. It worked pretty good. It detected a crack that was very difficult to see -- smaller than hair diameter. I removed the insulation of wire before attaching it to avoid any elongation or the covering layer due to crack.

But I will do a few more testing soon to find out how reliable they are.

:slight_smile:

16

Here is another aproach for a short section. How about thin drawn glass tubes? I am thinking about 1-2 mm in outer diameter. To use you would simply pressurize them with nitrogen and watch for the pressure to go away when the tube breaks with the crack. The glass could be easily made by a glass blower or Corning factory line. A tempered type glass is VERY fragile and the hardest thing would be installing it on the side of the concrete. This is just a basic idea as there would be a lot of details to think about.

Good Luck,
R

17

Sachindra:
Hello Arduino experts,

I need a help regarding my project of detecting a crack in concrete.

The project has two parallel concrete beams each 500 ft long, 18 inch wide, and 10.5 inch deep. As this is pretty long structure and exposed to the environment, it will have some hairline cracks in random spots. My aim is to detect where the crack occurred first and when did it occur in a small segment (25 ft) of the beam.

I am a civil engineer (but I have moderate skills in programming and biginner-moderate skills about electronics). I would appreciate if anyone could please help me find sensor or way to detect these hairline cracks (in the order of microns).

My initial ideas are:
Idea 1:
Using an arduino with multiple parallel connection. each of these parallel wires would then be attached to the side of the concrete beam (as shown in attached figure). Each arduino would measure the crack development in 25 ft section (with five 5 ft wire placed as seen in figure). In ideal case, after the beam has hairline crack, the wire would break and the circuit would open. From the location of that wire, I would know where crack occured and the arduino would have to record the time when that circuit opens.

One problem with this method might be,the wire might not break at all with that small crack.

Idea 2:
Using some acoustic sensor or infrared sensor or laser and measure the distance change. Sensor would be attached to the side of concrete with reflecting device/surface at some regular distance (as shown in attached figure). The sensor should then record the time when the crack occured by measuring the movement. This idea might have a lot of disturbances picked up as the beam is outside. In addition, the crack is so small (in order of micron) that the resolution of these sensor might not be good enough.

This is the first issue. The other things that arduino must do is record that time in SD memory card and transfer it through wifi. However, those would be just other standard shields being used.

I hope that I have made my problem clear. Please let me know if any more infromation is required. I really would appreciate any help.

Thanks,
S

Try this Glass breakage Alarm tape.

If you can seal it to the concrete is might work.

Chuck.

18

don't know if this is solved or not.
i figurre asking about concrete.
if that 24K carbon fiber is threaded through the concrete like a web from many cross directions, would that be better or worse than doing the same thing with metal (or maybe both)?