Appendix D 3

[FeSE]

Does anyone else view appendix D as a poor code?

Let's take a poll;
How many people use a simpler, or older anchor design routine?
How many people actually use Appendix D?

 

[PEinVA]

PackerFan,
Thanks for the information on the other forum, however, I disagree with you here. Hilti's HDA Undercut anchors is given a 1 for anchor category, as you mention above for least sensitive to installation. These bolts are all metric right now. If the contractor doesn't use the correct tool, bit and setting tool for this installation these bolts are useless. I was on a job where the Hilti representative (field engineer) came out to review the installation due to concerns the contractor would do it incorrectly, both on EOR's end and Hilti's.
What company is going to put out a bolt that isn't in category 1?

It doesn't seem to make sense to me but thats just my opinion.

RC
All that is necessary for the triumph of evil is that good men do nothing.
Edmund Burke

 

[ron9876]

I feel not having problems over a long period of time is very significant. In my opinion this is another case of the undue influence of out of control academics on the code process.

There are several things that are not exact models of physical reality but have worked for an extended time such as the equivalent frame method for slabs. I use the PCI method which is in their manual. At least it is presented in a usable manner.

 

[271828]

ron9876, I won't belabor the point beyond here because we've gone over it before and it's always a big fight.

The bottom line is that ANYBODY can come up with design provisions, be on committees, etc. If someone doesn't like how it's going, then he or she needs to step up to the plate and stop it. Otherwise, it's like someone who doesn't vote, but complains about what politicians do.

 

[JStephen]

"not having problems" can be deceptive. Anchor bolts that are designed for wind or seismic loading can be there for decades and never be loaded to their design loading. I do seem to recall a recent tunnel-roof failure due to dead loads on a bolt, though.

 

[ron9876]

Well it certainly is true that many structural elements never get fully loaded. But the precast industry has used headed studs to carry gravity loads for years. My point is that between engineer errors and contractor errors if there is something that we use that has a problem it will show up like corbel design has.

As far as code development is concerned there is no way that I have time or money to spend on committes. But it does seem to me that in about the last decade that we have had code development and revision at a runaway pace. Especially in the concrete code. For me it is really frustrating due to the lost time and money spent. I think Appendix D is an excellent example of a change for the sake of change. It was simple and it did work. Load testing indicated that the previous method worked fine so why buy new literature and try to fight thru a cumbersome (to be polite) design procedure when it wasn't needed.

 

[lkjh345]

ron9876:

Thank You for saying what most of the structural engineers I know really think about Appendix D!

 

[jt12]

RON9876 - I don't think that testing indicated the previous design worked fine. At minimum, the old method had a horrible correlation between expected and tested results. These are the reasons they tried to come up with a new method. I try very hard to avoid the app D method because it is cumbersome. Hopefully, it improves with code revisions.

 

[271828]

You make good points, ron. Hard to argue with most of them.

A little more about code development and academics, though:

The professors I know who are active on committees do this stuff almost as hobbies. They've studied a particular subject to death and are interested in it. They don't have time either, but do it anyway because it's fun and interesting. I really don't see why design professionals don't do this also. There's really very little work involved and meetings are only a couple of times per year in most cases.

For what it's worth, tenure and promotion are the ultimate goals of professors. Code provision adoption has *very* little impact on this. Dollar amount of externally funded research projects is the deciding factor by a very wide margin. They *say* that journal papers are also, but that's lip service. The universities want money.

I serve on one very significant committee and professors are only 20% of the group. They have an impact for sure, but they are far from overwhelming all the designers, fabricators, etc. in the room. I'm sure there are counter-examples out there.

Get in there and fight it out!

 

[271828]

I have an idea, in all seriousness. Not being sarcastic.

If some of you guys have studied App. D enough to judge it, then why not come up with a simplified version? Make some conservative assumptions, show when some limit states won't control, etc. I bet this method could be boiled down to something useful without a tremendous amount of effort. Write it up as a journal paper.

There's obviously enough interest and energy to get this done. Heck, maybe I'll do it myself, LOL. I'm actually a bit surprised that this hasn't been done already (perhaps it has).

 

[ron9876]

Guys as I said above check out how the PCI Handbook presents this information. It is similar but for me it is much easier to follow and use.

 

[lkjh345]

Attached is my version of a very workable method of selecting expansion anchors. It is used in a country where the structural code committees might actually allow practicing structural engineers to make a decent living.
For that reason, the identity of the country cannot be revealed.

This method has been used for years and years with nary a problem.

What do you think?

 

[FeSE]

Ha ha. Design tables are fine, but not fully parametric.

And, OK, Appendix D is not that bad, but it requires a program.

So, does anyone have an open source program to share?

 

[JStephen]

If you limit the scope of application, it's not too hard to come up with a spreadsheet for App. D. For example, if you always have the same bolt pattern and the same shape of concrete face, you can deal with it. But trying to generate a spreadsheet that works for any bolt layout whatever and any concrete shape whatever would be a bit of a trick.

In the case of the table above, you can generate something like that if you use the same bolts all the time, limit the spacing in various ways, etc., but I've never seen a table like that that would work for any bolt anywhere.

 

[vincentpa]

271828,
"No failures doesn't prove anything."
COME ON!!!!!!!!!!!!!!!!!!

The precast industry has not had buildings falling down around us everyday. That is the true test. A laboratory is only a simulation. Have you ever been to Rome. Maybe we should chastise the Romans for their building practices since they didn't have test reports before building say the Coloseum or the Pantheon. There is something to be said for experience. This is one arena where the Roman engineers would be laughing at us.

Another example is Pre-engineered metal buildings. They don't collapse from anchor failures. Go to an older pre-engineered building and look at the size of the pier the anchor rods are embedded into, usually 18x18 or less. That pier if design by App. D would be 2'-6" x 2'-6" or more. Come on!!!! The results from Appendix D do not match what we see in the field. Certainly, some of those structures are not seeing the full design load but I dare say most are! I have not reviewed how they are testing these anchors but it seems to me that the results are way too conservative.

Then the time it takes to calculate by hand...
Come on!!!!!!!

Go ahead and try to write a spreadsheet for ACI APP. D. You need a PhD in Excel. It would probably be better to write it in Visual Basic, patent it and sell it to all of us hapless engineers.

App. D comes from the Euro Code. Take one look at the EU's bloated constitution and it's not a far stretch as to why App. D looks the way it does.

 

[271828]

vincent, LOL on the hyperbole. These forums are too boring otherwise.

I'm not being sarcastic--totally serious: You and some of the other guys who feel very strongly about App. D. really need to get together and come up with a better way. I wholeheartedly believe it is possible.

One fast option would be to simply accept that the equations better predict the behavior, but are more complex than necessary. Then look critically at which limit states never control or only control in rare cases. Re-formulate the provisions into something easier to use, although probably a little conservative.

IF App. D is unreasonably conservative for some limit states, then figure out why and use another model or equation! This ain't rocket science.

Write it up as a simplified method and get it published in a journal.

"Don't have time, but academics do." is a sorry refutation of this idea, IMO. I've spent enough time around professors to see how the good ones do things like this. They do craploads of stuff like this on their own time. It's more like a hobby that spawns off some sponsored project that provides them with test results (but does not have the goal of Code provision development).

If there's not enough interest and energy to try and fix the problem, but IS enough interest to complain, then, well I don't have anything helpful to add (Mama said not to say anything if I couldn't say something nice, right?) -- LOL.

 

[ron9876]

The former PCI method which used the 45 degree theory is my recommendation. Performed just fine and didn't waste nearly as much time.

 

[271828]

ron, how do you conclude that it performed just fine?

Does anybody here really know WHY App. D was developed and adopted? Who sponsored the research? Why? Did some guy just come up with this stuff for the fun of it?

 

[whyun]

This program has limitations (single, two or four bolts only) but can be a useful "learning tool".

http://www.simpsonanchors.com/software/anchor_design.html

Hopefully engineers don't use this "tool" as a black box, until they get a full grasp of each governing case.

As much as I have my gripes about Appendix D, the concept is really quite simple. Get the factored sear and tension, get the governing tensile and shear strengths (using all the modification factors for spacing, edge, etc), then do the interaction.

 

[ron9876]

I read it in different publications. It is my understanding the new method was developed to be more consistent with test results (+/- 22 degrees vs. 45 degrees).

The new method IS much more cumbersome and IS time consuming. I think that the code development industry has lost sight of the best interest of the people that use the design codes. Especially the concrete codes.

 

[lkjh345]

ron9876:

Couldn't agree more!

Does anyone really use Appendix D on real structures? Or has everyone, like me, just developed simplified 'work arounds' that we know are conservative, but we use to get the work done without wasting too much time?