Direct tensile strength of concrete

[ajk1]

Does anyone have a formula for the direct tensile strength of concrete?

 

[jayrod12]

It's called the modulus of rupture

 

[Ron]

Well, sort of......
Modulus of rupture is done as a flexural test and does not directly measure the tensile strength of concrete. Another commonly used, but also indirect indicator of tensile strength is the splitting tensile test. Both will generally yield a tensile strength of around 10 percent of the compressive strength. Axial tensile strength of concrete is a more difficult test to run and generally yields slightly lower relational value to the compressive strength. One accepted relationship is

f(t)= 7.5 * ((f'c)^0.5)

 

[ajk1]

Yes I know those indirect formulae. Let me phrase the question differently. What would you expect the direct tensile strength of 35 MPa (5000 ± psi) concrete to be? If your answer is 10% or 500 psi, or 7.5 x 5000 [sup]0.5[/sup]= 530 psi, I am skeptical because we have never achieved anything like that when we due direct tensile tests of bonded topping and the fracture occurs within the parent concrete below. We would usually get 300 psi, perhaps 330 psi. I suppose we could do some direct tensile tests of monolithic concrete and see what we get...but I was hoping maybe those tests have been done and reported on somewhere. Any comments would be welcome.

 

[jayrod12]

Oh come on Ron, now you're just arguing semantics. whether the concrete is put into tension via bending or axial loads doesn't modify it's ability to resist said tensile stress.

 

[KootK]

Actually, according to code provisions, flexural tension and direct tension values are different. I got schooled on this myself a while back but can't locate the thread.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[jayrod12]

If he's going to tell me that it's 10% of the compressive stress for direct tension then I'm calling it the tomato-tomato argument.

 

[DamsInc]

Are these in-situ pulloff tests or lab direct tensile tests? What diameter is the sample?

 

[robyengIT]

in Italy (and I guess in Europe too) we use values as per attached table (NTC 2008 Standard ruling by law)

 

[Brad805]

The study below was to determine the early age strength of lifting hardware for precast concrete, but there is some testing that might be of interest to you. It also discusses the various methods used in different jurisdictions.

http://hosted.arinex.com.au/abstracts/cia2013/pdf/411310015Final00101.pdf

 

[Ron]

ajk1....comparing a properly done direct tension test in the lab to a "pull off" tension test of a concrete topping is apples and oranges. A pull-off test is subject to even more disturbance issues that a typical concrete core, so considering that, if you divide the pull-off strength by .67 or so (0.85 is used for cores in compression), you might get close to the actual axial tensile strength of the concrete.

A uniaxial tension test in the lab is done on a specimen that has no influencing finish issues common to floor slabs, even when some surface prep is applied before the topping. It is done on a prepared specimen, much the same as a compressive strength test.

A tensile bond strength of 300 psi is quite good!

jayrod12....no, not semantics....difference in stress orientation. In a flexural test you have both tension and compression acting on the specimen. In a direct tension test, there is no compression component or influence. In the splitting tensile test, Poisson's ratio comes into play as the test is actually done by compressing the sample along a straight line, not a defined area, and the failure occurs when the compression along that line causes the specimen to pull itself apart.

 

[racookpe1978]

Modulus of rupture is done as a flexural test and does not directly measure the tensile strength of concrete. Another commonly used, but also indirect indicator of tensile strength is the splitting tensile test. Both will generally yield a tensile strength of around 10 percent of the compressive strength. Axial tensile strength of concrete is a more difficult test to run and generally yields slightly lower relational value to the compressive strength

What is the structural strength of that nominal relationship over time? Would that formula be adequate for a 7-day concrete, or only after a 30 day cure time?

 

[dik]

Direct tension testing puts the entire cross section into tension and any weakness will precipitate failure, hence the lower value with direct tension. Flexural tension is more like the condition in service and only the outer fibre will be exposed to maximum tension. Splitting tensile strength, again only exposes a portion of the concrete to maximum tension, again, reflected in a higher tensile capacity.

For the Coef for determining flexural tension capacity, with airport pavements, the value is often as high as 9.0 to 9.2... but, a factor for fatigue, etc. is often included. Also quality control is often a little better.

Dik

 

[ajk1]

Thank you all.

My own understanding over the years has been pretty much what dik has said. I do believe that there is a significant difference between the direct tensile strength and other measures such as the modulus of rupture. Perhaps the splitting tensile strength may come closer to the direct tensile strength than the modulus of rupture.
Yes I did realize shortly after posting my comment about my experience with on-site direct tension test strengths, the point that Ron makes, namely that the quality of the concrete in the top few mm on a slab is not as good as below that, and so the direct tensile strength would probably be less than if that layer were removed before testing. However, our tests are done on projects where the floor was shotblast, so perhaps that weak layer was removed.

Since the tests we conduct on projects are of new topping on an old slab, and the results I am talking about are for those where the failure surface is in the old slab, the age of the concrete should not affect it.

Would be nice if there were a published paper somewhere that dealt with gave test results for lab direct tension tests on monolithic concrete. Maybe ACI has something or an academic has info.

The reason behind my original question was to determine if a structurally reinforced bonded topping to an old slab that will be shotblast to 5 mm amplitude, can be considered the same as if the concrete (topping and original slab) was all monolithic, and then horizontal shear stress at the bond line would not be a concern. The proposed live load on the composite slab is 250 psf (unfactored), so there may be some concern about the horizontal shear stress at the bond line and whether the bond is sufficient to safely resist the shear stress.

 

[DamsInc]

Just to be clear, you are saying "direct tensile test" but you are actually referring to pulloff tests. They are different. Direct tensile tests involve removing a sample and machining it in a way that it can be gripped and then pulled in the lab like a steel coupon, but apparently this test is very rarely done.

I suspect one reason why the pulloff tests are not comparable to the actual tensile strength is because they are typically done with a small (50mm) diameter disk which is cored into the substrate. Like Ron said, small diameter samples are more strongly influenced by minor flaws because they make up a greater proportion of the cross section. As well, the damaged surface area from the coring makes up a greater portion of the sample. You have more cut/disturbed aggregates which are no longer effective. Same principle applies for compression tests, which is why we use larger diameters for a more accurate measurement. You really can't compare this test to the actual tensile strength of the concrete. But you can use it to see where the weak link is: the new concrete, the old concrete, or the bond.

There is literature comparing the test results between direct tensile, splitting and MOR tests. I haven't read any, but my concrete textbooks do discuss this difference. It is established that MOR strength > splitting tension > direct tensile.

Maybe I'm not understanding the shear concern, but if the pulloff test is breaking in the substrate, than the bond and new concrete are both stronger than the original concrete, so why would there be a concern about that?

 

[cnorvell]

ACI 318-11 chapter 22 (plain concrete) gives you 5*(f'[sub]c[/sub])^0.5, which for 5000 psi concrete is about 353 psi, not to far off from your test results.

 

[JAE]

I think on a past bonded topping overlay we used 180 to 200 psi as a target goal on the bond tests...this was for 4,000 psi concrete over a roughened existing slab.

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[ajk1]

Thank you all again.

To DamsInc: the test procedure routinely used here in Canada is CSA A23.2 - 6B, "Method of Test to Determine Adhesion by Tensile Load" (I do not know if there is a corresponding ASTM test or not). The CSA test requires that the minimum diameter of the core bit shall be 3½ times the nominal maximum size aggregate size but in no case less than 75 mm". The inside diameter of the core is stated to be slightly larger than the fastening disk. I am perhaps a little surprised that you say that pull-off tests are normally done with 2" (50 mm) disks, as most concrete has 20 mm (3/4") diameter coarse aggregate. Anyway when we do the tests we use 95 mm diameter disks. I don't think that the 20 mm aggregate should then be a significant factor.

You do make a good point, namely that if the failure surface is in the original concerte of the substrate slab, then that shows that the bond line is as strong or stronger than the original concrete, irrespective of the failure stress.

I also appreciate that you have found that your reference book says that
MOR strength > splitting tension > direct tensile. Sometimes the good old fashioned medium is the best! That is what I had thought when I started this string. Can you tell me the reference book that you found that in...maybe we have it in our office library. Is it Troxell & Davis?

To Cnorvall; yes I know but that would be a permissible ultimate stress to be used in limit states design; that is not the actual ultimate failure stress. The actual ultimate would be some number of standard deviations above that.

To JAE; We would specify about 1.5 MPa (218 psi) minimum adhesion strength for normal garage concrete repairs. Using proper care in surface prep and surface cleanliness and a particular water based epoxy bonding agent with 24 open time, we can get 2.0 MPa (290 psi), but it does take care on site. However on the particular job currently under consideration, the live load is 12 kPa (250 psf) so we wanted something better than 1.5 MPa.

 

[Ron]

Racookpe1978....those relationships are based on f'c, which is usually considered the strength at 28 days.

 

[DamsInc]

To DamsInc: the test procedure routinely used here in Canada is CSA A23.2 - 6B, "Method of Test to Determine Adhesion by Tensile Load" (I do not know if there is a corresponding ASTM test or not). The CSA test requires that the minimum diameter of the core bit shall be 3½ times the nominal maximum size aggregate size but in no case less than 75 mm". The inside diameter of the core is stated to be slightly larger than the fastening disk. I am perhaps a little surprised that you say that pull-off tests are normally done with 2" (50 mm) disks, as most concrete has 20 mm (3/4") diameter coarse aggregate. Anyway when we do the tests we use 95 mm diameter disks. I don't think that the 20 mm aggregate should then be a significant factor.

Right, I wasn't sure what diameter you would be using, but 50mm would be the minimum. Which version of CSA 23.2 are you looking at? The 2009 version states that the minimum would be 60mm for the coring bit, for a 50mm sample (strange that they decided to revise that downward).

It was my interpretation that in CSA 23.2 6B, procedure 'A' was the pulloff test, and procedure 'B' was direct tensile, but I could be wrong on that. That's how I've seen/heard it described.

I also appreciate that you have found that your reference book says that
MOR strength > splitting tension > direct tensile. Sometimes the good old fashioned medium is the best! That is what I had thought when I started this string. Can you tell me the reference book that you found that in...maybe we have it in our office library. Is it Troxell & Davis?

From "Concrete Technology" by Neville & Brooks,