Crack Control

[JStructsteel]

If I have a beam that is going to crack, i.e. the stresses are more than sqrt 7.5F'c, does the ACI equation just control the crack for stress in the rebar, or does it also limit the visability of the crack? I konw its going to crack, but how big of a crack will it be. I thought it used to give you a crack width

 

[frv]

At service loads, cracks really should not be obviously visible. If you look closely at a beam, however, you will be able to see them.

Calculating the crack widths with any level of accuracy is nearly impossible (I'd venture to say you'd be lucky if your guess is within an order of magnitude of the actual width of the crack)

 

[Gumpmaster]

You can approximate crack widths with the ASD method in the ACI 350 appendix.

 

[JStructsteel]

thanks guys, I guess my ultimate dillema is if the spacing for the rebar is followed, does that mean the crack width is barely visable, or visable and not a considered failure?

 

[rapt]

jrisebo,

Nominal code rules for crack control normally try to limit the width of a crack to .3 - .4mm (about .01"). So you can see it but it is not a gaping crack.

This all assumes that you have allowed for all applied loadings in design (eg restraint to shortening etc).

 

[ahmedhegazi]

Actually crack width is variable according to the surface you want to control the crack width for, if yous structure is water retaining structure i go for 0.2mm, if it is liquid retaing with aggressive liquid , i go for 0.1mm

Normaly for crack width 0.2mm you will find that you use more steel than required for strenght

 

[Gumpmaster]

The ACI LRFD code doesn't really design, or limit cracks to a prescribed width. Crack width is actually really difficult to predict. It basically deals with how much strain is present at the extreme fiber. The more reinforcing present, the smaller the cracks will be and the closer they'll be spaced. Lesser ammounts of reinforcing result in bigger cracks at a larger spacing. The ACI provisions are written so that a steel tensile failure preceeds a concrete compression failure. A steel failure is much more predictable than a concrete failure, and steel yielding produces pretty good sized cracks. If the member has actually failed, the crack size should freak you out. Visible cracks doesn't necessarily mean that the member has failed. If you can stick your finger into a crack, it's probably failed.

The ACI ASD provisions limit the tensile stress in the steel based on how large of a crack you want to allow.

A good computer program to estimate crack widths is Response 2000 (search for it on Google).

 

[JStructsteel]

Thanks. I understand that ACI limits steel stress to 2/3Fy, so there is no chance of failure. Most designs that I have done are controlled by flexural stresses, not the crack control equation.

I am used to desgining uncracked concrete panels, and desgined a few stairs, and just wondered how much cracking would show.

 

[tuscan72]

what software is capable to carry out crack analysis of a concrete section?

Tom

 

[JKW05]

Refer to ACI-318 Section 10.6.4. Review of the older ACI codes back to 1995 will give some insight into the development of the code requirements in this section.

Before the '99 code, ACI prescribed calculation and limitation of "z" in kips/inch in an effort to limit objectionable cracking in beams and slabs. ACI 350-01 had the same method, with more strict requirements for "z".

 

[Lion06]

They changed this in the ACI code due to the inability to predict crack widths. Too many things go into predicting a crack width for it to be estimated even reasonably close. The old ACI codes that had a Z-cracking equation could not accurately predict crack widths and engineers had problems "meeting the code" since the crack widths were often greater than calculated.
The new provisions for spacing to control cracking eliminates this problem since as long as you meet the spacing requirements you "meet the code" for crack control.

 

[JKW05]

Not sure I agree with your assessment of the use of the "z"-factor.

The Pre-1999 use of this method was not to predict crack widths, but as a means of considering the distribution of the flexural reinforcement to minimize objectionable cracking, just as the current method is. In fact "z" was specifically defined as "quantity limiting distribution of flexural reinforecment"

With respect to "meeting the code", I do not feel there was a problem "meeting the code" using this method, as the ACI Code (Both 318 and 350) set specific upper limits for the resulting "z". Providing adequately distributed reinforcing to meet the code requirements was never a problem for me. The current method and the old method both rely on the accuracy to which fs can be predicted. The objective for the analyses are the same, the method has just been updated.

I will say, I found few engineers who actually took the time to check the z value. It may also be interesting to note that Enercalc 5.8 still calculates the z factor. Haven't got V6 yet.