Curious Question About Flitch Beam Design 7

[tc7]

I have an upcoming residential project that may require a stack of LVL's or perhaps a steel flitch beam configuration to replace a bearing wall spanning ~16 feet or so. To prepare, I've read as much of the available internet literature that I could find including the ubiquitous Destefano article in a 2007 Structure Magazine as well as a couple of good text books on structural wood design. I've also reviewed several dozen excellent Eng-Tips discussions on the subject of flitch beam design. With all this new found knowledge and my own familiarity with the NDS, I'm ready for the design work as soon as the owners are ready.

However, never in any of the discussions or references that I've seen has the effect of bolt holes been mentioned insofar as their reduction of beam section modulus and the consequent increase in the bending stress. A couple of 1/2" dia. bolt holes will take out a fair amount of cross section (~11% from a 1/2" steel plate). If those bolts are located ~2-inches from the top and bottom edges as depicted in the bolt pattern shown in the Destefano article, this could reduce the moment of inertia and section modulus of the steel plate by ~11% also.

This seems significant to me and perhaps should be accounted for in the design calculations, but I don't design flitch beams often (in fact, never). I would appreciate all thoughts on this matter and why the effect of bolt holes on bending stress and shear stress seem to be ignored.
Thank you.

 

[jerseyshore]

The reality is that on a home contractor installed flitch beam, there are probably 10 other things that would have to go right (or wrong) before the bolt holes changing the section modulus control. You'll be lucky enough to get the right sizes pieces out there and bolted together in any reasonable fashion before you got that close to failure due to net vs gross area.

 

[BridgeSmith]

There should be no noise in design work especially where public safety is concerned, just ask the structural steel folks - they've got their act together.

Maybe "noise" isn't the best term. How about material variability, unanticipated support conditions, unknown restraint conditions, etc.? The truth is that design in the real world includes alot of approximations, and because of that, we never really design 'that close to the edge'.

Rod Smith, P.E., The artist formerly known as HotRod10

 

[XR250]

"lost in the noise - Synonym lost in the underflow. This term is from signal processing, where signals of very small amplitude cannot be separated from low-intensity noise in the system. Though popular among hackers, it is not confined to hackerdom; physicists, engineers, astronomers, and statisticians all use it."

Basically what BridgeSmith said. I used to do signal processing work so it seemed like a good analogy to me.

 

[dauwerda]

If you're still viewing - I'm fascinated by your reply; I gather the steel Codes that you mentioned allow a member to exceed yield when analyzing a member based on "net area" i.e., cross section minus area of holes, but what is the allowance criteria? how far past yield does the steel Code permit

Keep in mind, steel yielding does not in and of itself constitute a failure. It really means two things: One, once stress goes above the yield point, the stress strain curve is no longer linear and strain will increase faster than it did in the elastic zone. Two, when/if the member is unloaded it won't fully rebound to it's original state (it will still recover the displacement that occurred in the elastic zone).

To simplify things, let's consider a tension member rather than a member in bending. If I have a 12in rod with a 1 inch cross section, a 50ksi yield strength and a 65ksi ultimate (or tensile) strength (this corresponds to A572 Gr. 50) and I load it in tension with 50 kips, it will be loaded to yield and would have a strain equal to P/(A*E), so 50kips /(1in[sup]2[/sup]*29000ksi) = 0.001724 in/in, so at 12" long it would stretch 0.0207".
At that point if we unload it it would return to it's original 12" length.

Now let's put ten 0.1" diameter hole in it every inch so that the cross section at those points is 0.9in[sup]2[/sup] and reload it with 50 kips. At each hole, the stress is now 50kips/0.9in[sup]2[/sup] = 55.6 ksi, this is past yield but it hasn't reached the ultimate strength of 65ksi yet so we know we don't have necking issues and it won't rupture. Now, I'm not sure what the strain is at 55.6 ksi, but if we assume it is double what it is at yield I bet that would get us close. So over the 10 holes we have 1" of length and that strain is 0.001724*2 = .003448in/in * 1 in = .003448. The rest of our rods length is 11" so it will stretch 11*.001724 = 0.019" for a total of 0.0224". If we unload the rod it will now have a total length of 12.0017". I would say this is not going to cause any issues in structural applications.

As far as the code goes, for tension members AISC 360 tells us to:
Check yield on the gross area with a phi factor = 0.9 (so 0.9*Fy*Ag)
Check rupture on the net section with a phi factor = 0.75 (so 0.75*Fu*An)
AISC also gives further clarification on calculating net section (bolts holes taken as 1/16" larger than the nominal dimension of the hole as well as other provisions for closely staggered holes). Again, the reason AISC allows this is as described above, localized yielding won't be detrimental to the member in any way.

For bending, AISC has the following provisions for holes:

 

[tc7]

dauwerda-
That explanation belongs in the Eng_Tips top ten most thorough and eloquent replies ever written (on any subject !). It erases all doubt in my mind about the prevailing design practice for steel flitch beams.

I'm certain the information presented in this thread by yourself, phamEng and ProgrammingPE is new-found and next-level knowledge to all flitch beam designers and explains away the noise that others have mentioned. This thread will probably become the new go to reference whenever the subject come up again.

Thank you very much for your generosity.


P.S. Your posts are so important to me that I printed them out and special tabbed them in my design notebook.

 

[dauwerda]

I'm glad you found it helpful, tc7

 

[bigmig]

I wish I had a job where I had the budget and schedule to care about holes in flitch plates.

 

[tc7]

That's what good engineers do bigmig, we worry about things we don't know. Ignorance can't be offset by cost or schedule. Consider yourself enriched, now you know why you don't have to care about holes in flitch plates.

 

[jerseyshore]

bigmig that's me on like 90% of these posts. But tc7 is right that a lot of the time we find out why we don't have to care about these things.

 

[BridgeSmith]

It's good to stop and think about these things sometimes, because occasionally you find something you realize that you should be caring about.

Rod Smith, P.E., The artist formerly known as HotRod10

 

[jerseyshore]

I agree with that. Especially with things that wouldn't be considered tried and true methods or conventional designs.

For me I'm always extra careful with things that don't have redundancy or can absolutely see the design loads. Balconies, decks, monopole signs, unbraced beams, sidewalk vaults, open pergolas, etc. You know things that can and do fall down.