Difference between designers fc' and as-tested fc' 2

[IJR]

You have done your calcs with 28-day conc. strength of say this much ksi. The building is put in place and later on tests on samples show a little deviation?

How much deviation is tolerable?

regs
IJR

 

[Ron]

IJR...ACI 318 provides the tolerance for compressive strengths, assuming sampling and testing at time of placement. Essentially, single test "outlyers" can be about 10 percent low, assuming the averages are adequate and enough tests are done. It is a little more complicated in that the supplier is supposed to prove his mix capability with enough data for statistical comparison (ACI 214). If these data cannot be produced then the design mix must be stronger to compensate for the lack of predictive data.

If testing the "in-place" concrete by coring, the allowance is that if the core strength achieves 85 percent of the design strength, then it is generally acceptable for strength purposes.

 

[IJR]

Thanx Ron boss and congrats for mastering the week

respects
IJR

 

[Ron]

IJR...thanks. Where are you located?

 

[IJR]

born in Zanzibar and currently working for a small but very active engineering firm in Ankara.

And one day you will be invited to visit wherever I will be then. Just allow some time.

Keep up the good work Ron

regs
IJR

 

[capper]

Gentlemen,
This is where I think permissable stress design falls down, and where limit state (or LRFD, I think) comes into its element. Full statistical data can be made available from the supplier, you could then match this with your design strength characteristic curve - and usually it is okay(ish!)

Colin Caprani

 

[emmgjld]

I don't want to get into a discussion about structural design methods BUT, CAPPER suggests that 'Full statistical data be made available from the supplier...' I daily deal with a supplier who cannot produce reliable or believable data and I have another supplier who has changed quality in the last 1 to 2 years and has yet to stabilize. Add to this occ. erratic testing quality and what numbers do you belive?

The answer to the original question by IJR requires some knowledge of the local suppliers, testers and the contractors. Full statistical information can be invaluable but, is often not availible or sometimes, believable. I think some consideration of the history of the local suppliers, testers and the contractors will be very important in the final decisions.

I say all this while being in the business and having worked hard to maintain good, competant personnel and a believable laboratory.

 

[Ron]

Emmgjld..I agree completely. Having been at this for longer than I care to admit, I have seen great difficulty in getting appropriate backup data. You would think it routine, but it isn't.

The good news is that it is getting better, at least here in the States. Automated batching and computerization have significantly improved the POTENTIAL to acquire good data, and to be able to apply these data to evaluative purposes. Keep in mind that ready-mix suppliers rely on multiple laboratories for their data in the normal course of business. One must assume that data from multiple laboratories are comparable. Not always a good assumption.