Anchor Bolt Design ASD vs. LRFD

[cdi12]

Catalogs for adhesive or mechanical anchor bolts usually give allowable and ultimate loads and their ratio is around 1:4. However ratio between nominal loads (using allowable stress design) and factored loads (using strength design) is less than 1:2. This means allowable stress design gives you anchor bolts about 2 times smaller than strength design. Am I missing something?
Thanks,
IV

 

[271828]

The LRFD resistance factor and the ASD factor of safety should be adjusted so that they give a similar result for some ratio of dead to live loads. What resistance factor is being used?

 

[Lion06]

Also, I personally would never use the Ultimate capacities (for adhesive anchors) even with factored loads. I would still apply a reduction factor to the ultimate load, just a smaller one if I am using factored loads and a larger one if I am using service loads.

 

[Galambos]

we had a representative from Simpson give a lunch-n-learn. He clearly stated that we should NEVER use the ultimate loads nor apply a reduction factor to them to use with factored loads.

 

[271828]

Galambos, I have no clue whatsoever how the Simpson guy could get that.

It's entirely possible to have the same reliability using factored loads and a resistance factor as with unfactored loads and a FS.

Care to elaborate? That's one of the nuttiest things I've heard in a while.

 

[cdi12]

Thanks to all for your valuable inputs. You have already noticed I made a typo - ASD gives you LARGER bolts than LRFD not smaller. For example ASD load comb is 1,000# tension per bolt and LRFD 2,000#. 1/4” dia bolt w/ 2” embedment has allowable load 560# and ultimate load 2100#. Therefore using LRFD this bolt is O.K. but using ASD is not – bigger bolt has to be used. 271828 I thought Resistance Factor for LRFD and Factor of Safety for ASD are already included in anchor bolt load tables.
IV

 

[Lion06]

NO, OH MY GOODNESS NO!! The ultimate values listed in Hilti or Powers for the adhesive anchors are the actual ULTIMATE values. They place a F.S. of 4 (or close to it in most cases) for allowable loads. I use a much higher F.S. myself if it is a critical building component, I will possibly let a smaller one slide for a less important component.
The bottom line is that the ultimate loads listed in the catalog have no reduction factor placed on them. That is why you are getting such a disparity is your bolt sizes.

 

[271828]

StrlEIT, but you can set a resistance factor to use with factored load combos and accomplish the same overall FS, right? Maybe phi=0.3 or something similar.

 

[Lion06]

271828-
Correct. I would think the method is secondary and the result is primary. I guess you could probably double your Load Factors and then actually use the ultimate value and accomplish the same thing. I was just trying to point out that using traditional Load Factors you should absolutely not use the ultimate values with no reduction.

 

[PackerFan]

The ultimate loads derived from testing using the ASD design methods currently published by an mfg (Hilti, Simpson, etc.) should not be used with any pre-existing phi factors since there has not been any testing in correlation with what that Phi factor should be. This is one of the reasons that the methods of ACI 318 App. D came into existence. Phi factors for each anchor are derived from how well the anchor performs in many conditions (cracked concrete, incorrect installation, wet holes, core drilled holes, etc.) Many products from different mfg's have very different standard deviations on the testing values; therefore a universal phi factor would not be possible. There is no possible way to derive LRFD loads from the ultimate loads for most products since this phi factor is unkown. Only products that have been tested to the new standards will give you the appropriate phi factors. Design data for LRFD design will be presented in each mfg’s ICC report. There are only about 5 mechanical anchors that have this testing completed.

 

[271828]

StrlEIT, I'm with you, but we should clarify. You'd never change LOAD combination factors to account for RESISTANCE variability. You'd use a smaller resistance factor and keep load factors the same, right? The load doesn't know what kind of anchor you're using, so it won't be any more variable.

This is just like the new AISC Spec. Omega and Phi are set up to give the same overall factor of safety for a specific L/D load ratio.

For anchors, one could use this same L/D ratio and the given ASD FS and derive an equivalent phi factor. It's probably be very small, like 0.3, but I'm too lazy to calc it tonight, LOL.

 

[Lion06]

271828-
I agree. I'll calc it on monday! lol

 

[UcfSE]

If you choose to use LRFD with manufactured anchors, you should use the values calibrated for these. The manufacturer may be able to recommend something. Don't just make up your own values, and don't use values not intended for your application, and don't modify the load factors. That's not engineering. Resistance factors will be based on the standard deviation of the test values, and the ultimate load on the 5% fractile of the test results, not the average, not the lowest and not the mean or median.

There's nothing wrong with using LRFD, but like anything else it must be done properly, not half-way and not unreasonably over or under conservative. In fact you will have a better understanding of your reliability using LRFD than you would get with ASD. There is also an ACI publication, 355, which discusses post-installed anchors and their qualifications if you are interested in some bathroom reading.

 

[271828]

UcfSE, I agree and would certainly just use the published allowable loads and sleep ok at night.

As for using factored loads with appropriate resistance factors, I was just typing philosophically. One should be able to set the resistance factor to get the same overall factor of safety.