Ultra high strength concrete / mortar

[pythonex]

I hope this is the right forum to post in...

(For Research) : I'm trying to do a concrete mix (mortar for now before adding fibers) with strength above 18 ksi (120 MPa) at 28 days. So at 7 days I need at least 15 ksi. This is known as ultra high performance concrete and many researchers even achieved much higher than this with W/CM ratio as low as 0.15.

However I cant seem to get beyond 10 ksi at 7 days (which wouldn't reach 18 at 28 days). The major problem I see is that when the 2 inch cube specimen are tested, the core is still powderous. It is not uncommon to find unhydrated particles under very low w/c ratio, but why is it all in the core? It is as if the cube faces are setting fast and the core remains powder. I tried W/CM ratios of 0.25 , 0.225 and 0.2. Lowering the ratio doesnt seem to affect the strength.

Do I need to use retarder? I'm using a food mixer to prepare small batches. Is it over mixing? or need to mix for more duration?

Below is the mix proportions I reached this strength with, which are quite similar to proportions used in many successful researches done by others.

Cement 1200 lbs
Silica Fume 300 lbs
Find Sand (less than 600 microns) 2190 lbs
Water 337 lbs
polycarboxylate-based HRWR (50 lbs or as required)
(W/CM = 22.5 % )

Thank you!

 

[hokie66]

Too much research. We don't need 120 MPa concrete.

 

[EngineeringEric]

You need to use a fair amount plasticizer for that low W/C and then I would say over mixing is not an issue, but i do not recommend a food mixer. In Grad school i played with some .10-.15 W/C and never had any 'dust' in the center. The concrete paste should be well mixed before you form specimens, and compacted after forming.

As for the hydration of the cement and the formation of your matrix-structure and overall hardening.. With low W/C i would believe it has to do with your drying humidity. If the environment is not humid enough your samples may be drying to fast and losing all that water before any reactions can happen, so you have a true final w/c of say 5% all in the surface (?)

Also, 2" cube seems small for research, why not 6" cylinders and powerful presses

 

[TXStructural]

Make sure the aggregates are saturated surface dry (SSD). Dry aggregates will extract moisture from the paste and dessicate the cement.
W/CM below 0.2-0.35 is generally useless, since you need to have sufficient moisture to hydrate the cement. (0.26 is not necessarily an accurate number for full hydration.)
Mix the concrete long enough to get everything wet and start the reactions (2-3 minutes.) If the sand is bone dry, mix sand and water and let it soak for a few minutes before adding the portland and silica fume. Also keep in mind that silica fume usually makes a mix sticky because of its excessive fineness (making the suction between particles greater because there are lots of narrow spaces between particles and less relatively larger space.) It takes a fair amount of water just to coat the surface of all particles, and you have lots of fine particles.
And 2" mortar cubes are not a fair representation of concrete strength, else we would all use them.

 

[cvg]

for high early strength and low w/c ratio you will need the finest grind on your cement that you can get. not all portland cement is created equal. aggregate quality is also essential and generally you will get lower strength from smaller aggregates such as sand and from weaker minerals.

 

[pythonex]

EngineeringEric, TXStructural and cvg, Thank you for your contribution.

I'm not using cylinders because their surface would require end grinding (capping doesn't work with high strength), and the end-grinder is not available.

I will use more amount of superplasticizer, however I'm not sure what to do regarding humidity. I'm sealing the cubes with impermeable plastic plates pressed on the whole mold. Is this enough or should I transfer the molds to the moisture room as soon as the surface dries?

I tried using accelerator but didn't help with the hydration of the core. Maybe I'll consider a retarder.

Finally, my sand is air dry and I'm adding the extra water equal to the absorption percentage, do you still recommend wetting the sand with the amount of water specified for my mix?

Thanks again for your replies.

 

[EngineeringEric]

I used caps back in the day for cylinders, we would epoxy a cap to the top and bottom and use that as our surface, it was critical to get it all level but then again a little local failure only messes with he data not the results. And i believe cubes have issues with the corners providing additional capacity to the sample. I cannot recall the details but i believe there is a conversion to go from cube to cylinder but a cube that small is difficult to correlate.

When using a retarder i would be curious as to the strength gain over time, your 7day of conc w/o may have more capacity then one w/, but the 28 day be opposite.

Also, when you are working with capacities in excess of 7000 psi, the cement paste is no longer the governing failure mode, the aggregate probably is. As cvg mentioned. when using real aggregates (larger stone) the stones split/shear and it is along dense packed lines of stone that the exploding failure will happen, this is very different than the normal failure of a 4000psi mixture.
Much louder, more exciting, and don't forget the shrapnel!

And you have to measure the water that you added to wet your dry sand, the w/c ratio is water for total mixture no matter when the water is added.

 

[concretemasonry]

After the cubes are broken, what will they be compared to to get a relationship to real concrete? A 2x2 mortar cube fails in a different manner that most concrete does.

A sample with a different aspect ratio of 2:1 would be better to use to measure relationships. There are smaller cylinders than the 6x12 that have been used in many tests because of the availability of testing equipment. Granted, 2x2 cubes are easy to work with because of the size, but the results are very difficult to quantify.

The 2x2 samples are satisfactory for mortar for research because the are closer to the real failure of mortar. Mortar strength is not a large factor in masonry, so no field tests are used.

If you can get into the realm of a 2:1 (height/width) relationship, you can get some historic correlation. Also, The gradation of the aggregates are critical to maximize the strength since one size aggregates (especially if rounded) provide lower strengths.

If this is just for project not relating to actual concrete, the 2" cubes may work for procedures, but not actual strengths.

Dick

Engineer and international traveler interested in construction techniques, problems and proper design.

 

[pythonex]

Thanks all for the input. One paper about the same topic goes in detail about strength comparison between cubes of different sizes to cylinders of different sizes too. I was just concerned about the state of the inner core of the specimens but I'll try some of the suggestions above and see what changes.

 

[darthsoilsguy2]

cubes might be most appropriate. depends on your application. i have a hard time imagining an application to use this in ftgs, slabs, walls, or columns as a primary material. If you were thinking building concrete, cylinders are the way to go.

there are a lot of high-end mortar/grout blends in the market.... latex, ceramic, and portland cement based. i think most of those run cubes. I believe the applications you would be pursuing would likely be appropriate to cube-test.