ACI 318-11 R/F Strain Limits per 10.3.5

[dmj2015]

I am interested in people's thoughts on designing non-prestressed r/f concrete members using a steel strain of .002 with a corresponding phi value of 0.65 given the requirement of Section 10.3.5 which appears to place a lower limit on the strain to 0.004? Thanks for the thoughts.

 

[rapt]

Why do you want to use a steel strain of .002. That is the yield strain.

If you design to that you have a balanced section, but ACI has always required limiting the steel ratio to .75Pbal and in later times possibly even less, like the Australian code which limits it to .67Pbal.

This is the equivalent to the old neutral axis depth limitations and is independent of steel strength.

Design to a steel strain of .002 has never been allowed or recommended as the section is essentially over-reinforced and subject to compression failure.

 

[dmj2015]

Thanks for the thoughts. I agree, designing to a strain of .002 is probably not a good idea, my question is why does it seem to be allowed per Section 10.3.3 of ACI 318-11?

 

[JoelTXCive]

There are probably better ways of stating this, but...

My understanding is that if you design your members so that the strain in the steel is .005, then ACI allows your capacity to be 90% of the nominal capacity(phi factor = .90).

You can design to much lower strain values down to .002 if you choose too, but the penalty is a reduced phi factor. You are only allowed to hit 65% of nominal capacity (phi factor =.65) when you don't have much strain.

I thought the roots of this are safety concerns. In a failure situation, A steel strain of .005 will result in large deflections and be visually noticeable. People will have time to get out of the building or structure before the concrete fails on the compression side (.003). The ACI 'rewards' this type nice ductile failure mode by allowing you to get closer to the nominal capacity.

On the other hand, if you only see .002 steel strain prior to failure, then there will not be large deflections and problems will not be visually noticeable prior to failure. This type of design is 'punished' and you are not allowed to get as close to nominal capacity.

When the steel fails on the tension side it is nice, pretty and slow; when the concrete fails in compression it is violent and sudden.

 

[JoshPlumSE]

My impression was moving from a 0.75 rho balanced to a strain based phi factor was a was of logically bridging the gap between the phi factors used for beams and columns.

There are times when you have "column members" without as much compression or flexural members that do have significant compression. Therefore, using a phi value between the two extremes makes some sense.

Granted, I'm not a PT guy and really haven't thought much this concept can (or can't) be extended to pre or post tensioned beams. Rapt certainly should be more of an authority on that subject.

 

[rapt]

I think the change to the strain logic was more to be consistent with different strength steels. A fixed neutral axis depth limit was based on a certain steel with certain stress/strain properties. Changing to a strain limit gets away from that limitation.

The 2011 code also had a limit for PT members (.005 I think). This seems to have been lost in the 2014 code, as I have discussed privately previously with Ingenuity. I hope that this was a mistake in the reorganisation of ACI for that version as the current 2104 code seems to have no limitation on ductility for PT members (or they have moved it somewhere that I cannot find).

10.3.3 is the general rule covering the basic logic for all members. defining tension controlled and compression controlled.
Columns can be either tension, intermediate or compression controlled and the phi factor will adjust to accommodate this.

The main logic is that flexural members should be tension controlled, so 10.3.5 controls this. It is not logical to design a flexural member as compression controlled as then the steel stress will not reach yield. So normal design approximations cannot be used and strains have to be determined to determine the steel stress at ultimate.