I am trying to create load~elongation, stress strain, stress~CMOD relation of concrete prisms in tensions. I have done preliminary 4-point flexural tests on RC prisms, the prisms fail suddenly and the declining branch after crack formation cannot be achieved. I will really appreciate it if someone can guide me in this regard. The machine I am using is Instron 100KN and the displacement controlled loading rate i used was 0.06mm/min.
Does inverting the supports and loading points make a difference? I have seen the inverted arrangement (loading points bottom, supports top) in the testing scenarios for plain concrete prisms. Does it make a difference?
I shall really appreciate your help in this regard.
~usman
Dear Usman,
I know it is hard to set a test to observe declining curve. I had similar problem with uniaxial tensile test for concrete cylinders. Even very low loading rate could not achive it. In my test I used notched (see figure) with specimens.
The solution for this problem was to put a gauge (LVDT) which take measures of displacement just at a notch (the base of notch was about 40 mm) while notch has about 5 mm. And the control of displacement was made by this gauge and the loading rate was connected with this gauge next at notch not with displacement of the supports.
The interesting thing was that the crack at notch of course increase even with lowering the stress (declining branch), but the the distance between supports of decrease (not increasing like the crack).
I think it could be difficult if you want do this without the notch. But maybe you can use same gauge in the region (between two loading points, maybe with some margin outside). That gauge should control the loading rate.
I also think that it can be difficult to find gauge which is able to communicate with a testing machine, that the loading rate is controlled by it not by the device which do this in normal situations. Sometimes it needs some spacial programming/instructions for testing machine to a gauge as a rate control device.
If I don't write it clear please contact with me.
Dear Mr. Sokal
Thanks for your response. I have already cast the specimens and they dont have any notch. The arrangement is here. The only option I have is to control the displacement. I want to observe the crack formation slowly. But what you have said, using the output of LVDT as control for loading rate, is something new. I will look into it. If there is some positive development, I will share.
Thank you for your response.
~usman
Since the crack formation, propagation and ultimate failure are very close it is normal to have difficulties with the descending branch. Did you perform test planning and simulation of the 4-point test? If you do a good quality simulation (which is possible since you already have the tensile strength and the initial modulus) you'll find out how slow you need to increase the displacement in order to get the curve. In my opinion the use of LVDT proposed by Mr Sokal is reasonable. You may also try to create notches but you should be careful to not destroy the samples.
Good luck!
@Mr. Andreev
Yes I checked the machine, I have ordered the lvdt for connecting to the machine for strain controlled testing. As soon as it arrives, I will proceed with the testing. I will respond here again with the outcome.
I tried it to the slowest in the machine as 4 point bend test, it has brittle failure. Which leaves us to the LVDT method as proposed by Mr. Sokal.
Thank you very much for your response.
I'll keep in touch.
Have a nice day.
~usman
- Badges
- Science topic
- Similar topics
- Engineering
- Civil Engineering