There are many shear models for reinforced concrete elements without shear reinforcement and each model shows an uncertainty in shear strength prediction. Other approaches like artificial neural network (ANN) are also used in shear strength calculation. The ratio between experimental and calculated shear strengths teta = Vexp/Vcal is often used to analyse the model uncertainty.
To avoid a failure in the evaluation of uncertainty of any shear model, I would like to know what is the minimum value of variation coefficient of teta (= Vexp/Vcal), firstly for normal concrete beams. is teta = 0.05, 0.10 oder 0.15? How to determine this value?
Best regards,
Ngoc Linh Tran
Dear Ngoc,
I feel you need to evaluate the correct shear modulus of concrete by conducting the laboratory test on reinforced concrete element and use those values for numerical analysis. I am sure you will get accurate matching by using the correct values of shear modulus..all the best..
Dear Mr. Karthigeyan,
thank you for the reply.
In numerical simulation, the shear modulus of cracked concrete is often taken into account by shear retention factor. I aggree with you that basic tests are always required (and very important) to determine the behaviour of material. However, the implementation in the calculation (also in numerical analysis such as with FEM) is not always correct (as it should be) due to different opinions and understandings and of course due to the complexity of the behaviour of cracked concrete.
Actually, I am only interested in how to estimate the minimum value of variation coefficient of ratio teta = Vexp/Vcal, maybe independent on a specific shear model (with assumption that the pure model uncertainty is zero). By chance, I have read a paper, in which ANN was used and a variation coefficient of teta was found as 0.05 for model uncertainty (with large database). I dont believe this value and therefore there is the question above.
Regards,
Linh
Mr.Ngoc,
i agree to your point that FEM may give incorrect results. I also suggest you to do the reliability analysis to get the correct variation coefficient..hope you aware all these things..all the best..
Hi Ngok,
I believe that you are after a strength ratio and not a modui ratio. Please confirm.
You need therefore to define your
Vexp and Vcalc.
Also, to elaborate on how Vexp is calculated from the experimental results and what is the state of stress therein.
The same applies to the Vcal.
In other words, structural information of the specific case need to be provided. Also, the local state of stress where you consider your ratio.
Kind regards,
Hesham Elshazly
Hi Hesham,
the shear strength I mentioned here is the shear capacity or shear resistance of a reinforced concrete member, unit in kN. Vexp is the experimental shear capacity (determined by a shear test). Vcal is the calculated shear capacity using a shear model, which can be empricial or based on mechanic background, can also be analytically or numerically formulated.
if I know only concrete compressive strength fc, as presented in many shear test databases for reinforced concrete elements, how large is the MINIMUM coefficient of variation of ratio between Vexp and Vcal ?
The idea of the question is that if we know the minimum COV of Vexp/Vcal for a given problem, such as shear capacity of concrete members without shear reinforcement, we can say a certain shear model is good for design and no further improvement is required (when we already define the input parameters for calculation).
Regards,
Linh
Hi Ngoc,
maybe this paper in Structural Concrete can answer your question:
Article Uncertainty in shear resistance models of reinforced concret...
Best regards
Hi Patrick,
thanks for the paper. In the paper, a mean value of 0.91 and an associated coefficient of variation of 0.14 were found by evaluating the shear equation without shear reinforcement according to fib MC2010 Level II with 184 beams.
The sample size is small (184 beams). Therefore, COV of teta can be larger if another database is used for the evaluation. The result is thus not good enough to give a general conclusion.
What I am interrested is the (theoretical) MIMINUM coefficient of variation of teta when we use a very large database, e.g. 1000 beams, to evaluate a shear model.
Regards,
Linh
- Badges
- Science topic