Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
Gusset Plate in Compression 4
- Thread starter formi
- Start date
-
1
- #2
CELinOttawa
Structural
To which code? That matters quite a bit...
NZS 3404: HERA BCRs (Available from HERA for a modest fee)
AISC: Design Examples (Huge document available online for free)
Also can contact:
- US & Canada: Steel Solutions Centre
- UK: Corus Steel Support
NZS 3404: HERA BCRs (Available from HERA for a modest fee)
AISC: Design Examples (Huge document available online for free)
Also can contact:
- US & Canada: Steel Solutions Centre
- UK: Corus Steel Support
- Thread starter
- #3
Actually, I design it to Eurocodes but there is no procedures for that kind of connx, so I'm looking for any other sources. In AISC Design Examples (13th Edition) I found example only with diagonal in tension, not in compression what makes a big difference...
-
2
- #4
This might be of help
CELinOttawa
Structural
What a find Desertfox! A star for you!
That said, the OP should not apply CFS methods to regular steel design...
That said, the OP should not apply CFS methods to regular steel design...
-
1
- #6
The AISC Seismic Manual covers this connection well. I'm pretty sure that Akbar Tamboli's steel connection text does too. Googling the uniform force method is likely to turn up a bunch of good articles by Thornton & Muir as well.
You'll want to have a black belt level of comfort with the Whitmore Section: Link.
@CEL: while I've never before thought to attempt it, I think that one could conservatively design hot rolled steel using CFM provisions. The reverse is untrue of course. Thinner steel just requires more attention to plate buckling and effective section properties as far as I can tell.
The greatest trick that bond stress ever pulled was convincing the world it didn't exist.
You'll want to have a black belt level of comfort with the Whitmore Section: Link.
@CEL: while I've never before thought to attempt it, I think that one could conservatively design hot rolled steel using CFM provisions. The reverse is untrue of course. Thinner steel just requires more attention to plate buckling and effective section properties as far as I can tell.
The greatest trick that bond stress ever pulled was convincing the world it didn't exist.
Residual stresses and cold work strain hardening would need to be dealt with appropriately I suppose... I take it back, I'd be careful using CFM for hot rolled design in general. For straight up plate design, things should translate fairly easily though.
The greatest trick that bond stress ever pulled was convincing the world it didn't exist.
The greatest trick that bond stress ever pulled was convincing the world it didn't exist.
- Thread starter
- #8
Thanks KootK! That's what I was looking for!
I have one more question guys: why w-shape diagonal braces in braced frames sometimes are oriented "normally" and sometimes are rotated 90 degrees along its axis?? Is it connected with buckling lenght or any seismic provisions? I use buckling lenght factor 1.0 for both: in-plane and out-of-plane buckling but I'm not familiar with other codes than EC's.
I have one more question guys: why w-shape diagonal braces in braced frames sometimes are oriented "normally" and sometimes are rotated 90 degrees along its axis?? Is it connected with buckling lenght or any seismic provisions? I use buckling lenght factor 1.0 for both: in-plane and out-of-plane buckling but I'm not familiar with other codes than EC's.
The orientation where the brace web is aligned with the gusset plate seems the most direct and rational to me (less tension lag etc). I'm not sure what advantages the other orientation would have. Maybe project specific spatial issues? Your plastic hinging is usually meant to occur within the gusset plate so I can think of no seismic benefits.
I did a project last year that involved heavy steel bracing. The fabricator elected to go with mated base plates: one on the brace and another on the gusset, with stiffeners. It went together pretty smoothly. That was a compression only application however.
The greatest trick that bond stress ever pulled was convincing the world it didn't exist.
I did a project last year that involved heavy steel bracing. The fabricator elected to go with mated base plates: one on the brace and another on the gusset, with stiffeners. It went together pretty smoothly. That was a compression only application however.
The greatest trick that bond stress ever pulled was convincing the world it didn't exist.
CELinOttawa
Structural
Well that was disappointing... I had quite the diatribe about assumptions and appropriateness of methods forming in my head as I read your first post Kootk...
I am both happy and pointedly saddened that you withdrew your point of view. *evil smiles*
@molten: No worries... Well deserved praise. One of the more useful posts I have seen lately!
I am both happy and pointedly saddened that you withdrew your point of view. *evil smiles*
@molten: No worries... Well deserved praise. One of the more useful posts I have seen lately!
CELinOttawa
Structural
Lol... That should have been @DesertFox... I have no idea why I said molten... Working on a Sunday. *sigh*
@CEL: if you've already got your debate ducks in a row, I'd still love to hear your reasoning for why CFM provisions couldn't conservatively be applied to rolled steel. While I didn't want to give a blank check to anyone who might read this thread in the future, I still mostly stand by my original comment.
The only factor that I can think of that would make hot rolled design by CFM methods potentially un-conservative would be the strain hardening bit. And I would think that a thoughtful practitioner could get around that by:
1) Not taking advantage of it in the CFM methods or;
2) Recognizing that strain hardening probably improves thicker elements more so than thinner ones.
There... I've reopened Pandora's box and invited you out to play.
The greatest trick that bond stress ever pulled was convincing the world it didn't exist.
The only factor that I can think of that would make hot rolled design by CFM methods potentially un-conservative would be the strain hardening bit. And I would think that a thoughtful practitioner could get around that by:
1) Not taking advantage of it in the CFM methods or;
2) Recognizing that strain hardening probably improves thicker elements more so than thinner ones.
There... I've reopened Pandora's box and invited you out to play.
The greatest trick that bond stress ever pulled was convincing the world it didn't exist.
CELinOttawa
Structural
Well alright then.... Let's set aside the issue of additional strength in a small cross section since, as you say, that can easily be ignored. They also apply most directly to the calculation of the section strength, so can be easily isolated and grouped for "avoid".
Here are five reasons off the top of my head:
- The failure modes of materials are highly dependent upon the exact material properties and St. Venant's principle (local versus global effects in this context). When the steel strength increased by 50 MPa, for example, a number of sections had their strength decrease. That's because the governing failure mode shifted, and the additional strength brought about (theoretically) an earlier failure. This means that formulae adjusted for one set of yield strength should not be used for another. Let's assume you are dealing with base steel within the AISI (or equivalent AS/NZS 4600, EC, etc) tolerances and continue...
- Limit states design material and resistance factors are again tuned to material properties and quality control set against statistical information for the intended material in a given range. Thicker materials are far less predictable, and the reduction factors in the CFS codes should not be applied to thicker steels. Let's say you recognize this problem, identify and sub out all the right reduction factors, and continue...
- The bearing and shearing failure modes for connections into CFS are fundamentally different than they are for rolled steel plate. CFS is almost never going to shear the fasteners, and as a result this is neglected as a form of failure in most designs, design examples, and even in some codes. Thin steel plate simply does not shear the fasteners; the materials builds up, crumples, and the failure is more akin to excess rotation and other "undesirable" boundry conditions unless you push the test specimen/building to an extreme. I was the forensic engineer on a CFS structure which collapsed under snow load, and many of the connections were still physically together, just twisted like my own tin-foil hat. Let's say you add in the required bolt shearing checks and continue...
- The shear block tear out are insufficiently detailed and resemble the previous generation of s-distances, etc. In this case you're probably safe, but throwing out the latest research.
- You need to recognize, pick through, and potentially may accidentally discard or not adjust stability formulae which are not easily sorted through. Effectively LTB and other concerns in CFS are compound formulae with multiple iterations and layers of calculations required. They also make little sense in the context of rolled steel sections - Combined Bending and Shear, or sorting through the use of elastic buckling load and section compressive load. Substitute, sort through, substitute again... It is simply asking for trouble.
The European Space Agency and NASA couldn't get a freeking satelite around Mars because they made a conversion mistake. It just seems like a bad idea to start on the wrong foot and then try to make a few dozen substitutions to get it right...
Here are five reasons off the top of my head:
- The failure modes of materials are highly dependent upon the exact material properties and St. Venant's principle (local versus global effects in this context). When the steel strength increased by 50 MPa, for example, a number of sections had their strength decrease. That's because the governing failure mode shifted, and the additional strength brought about (theoretically) an earlier failure. This means that formulae adjusted for one set of yield strength should not be used for another. Let's assume you are dealing with base steel within the AISI (or equivalent AS/NZS 4600, EC, etc) tolerances and continue...
- Limit states design material and resistance factors are again tuned to material properties and quality control set against statistical information for the intended material in a given range. Thicker materials are far less predictable, and the reduction factors in the CFS codes should not be applied to thicker steels. Let's say you recognize this problem, identify and sub out all the right reduction factors, and continue...
- The bearing and shearing failure modes for connections into CFS are fundamentally different than they are for rolled steel plate. CFS is almost never going to shear the fasteners, and as a result this is neglected as a form of failure in most designs, design examples, and even in some codes. Thin steel plate simply does not shear the fasteners; the materials builds up, crumples, and the failure is more akin to excess rotation and other "undesirable" boundry conditions unless you push the test specimen/building to an extreme. I was the forensic engineer on a CFS structure which collapsed under snow load, and many of the connections were still physically together, just twisted like my own tin-foil hat. Let's say you add in the required bolt shearing checks and continue...
- The shear block tear out are insufficiently detailed and resemble the previous generation of s-distances, etc. In this case you're probably safe, but throwing out the latest research.
- You need to recognize, pick through, and potentially may accidentally discard or not adjust stability formulae which are not easily sorted through. Effectively LTB and other concerns in CFS are compound formulae with multiple iterations and layers of calculations required. They also make little sense in the context of rolled steel sections - Combined Bending and Shear, or sorting through the use of elastic buckling load and section compressive load. Substitute, sort through, substitute again... It is simply asking for trouble.
The European Space Agency and NASA couldn't get a freeking satelite around Mars because they made a conversion mistake. It just seems like a bad idea to start on the wrong foot and then try to make a few dozen substitutions to get it right...
Thanks for that CEL. In my opinion, all of your most persuasive arguments have a common thread: the statistical calibrations built into our design procedures vary from one material to another.
I agree that the statistical calibrations matter. However, it's not as though we're comparing CMU and CFM here. I'd expect CFM and hot rolled steel to be quite similar. Unfortunately, I'm not in a position to track down all of the research that would be required for me to back up my opinion. Instead, I'll throw out the following:
1) S16-09 13.5.c.i specifically directs designers to use S136 for hot rolled sections with high slenderness components. Clearly, CISC thinks that the provisions for CFM and hot rolled steel are somewhat interchangeable. The suggested modification procedure is entirely manageable.
2) Way back when, I did some reading on the development of the ancient SSRC research that has informed our column design procedures ever since. I was shocked that all of the testing was done on very small sections (W8?). Meanwhile, folks are out there designing Mega-Columns for skyscrapers using those same "calibrations". I submit that, in all likelihood, the statistical variation within hot rolled steel design is probably as great as that between hot rolled and CFM.
3) I should make it clear that I would never propose using CFM procedures for hot rolled steel where analogous design procedures for hot rolled steel already exist. I'd only consider doing so if CFM code provisions/testing offered a solution to a design problem that AISC/CISC hadn't gotten around to addressing yet.
The greatest trick that bond stress ever pulled was convincing the world it didn't exist.
I agree that the statistical calibrations matter. However, it's not as though we're comparing CMU and CFM here. I'd expect CFM and hot rolled steel to be quite similar. Unfortunately, I'm not in a position to track down all of the research that would be required for me to back up my opinion. Instead, I'll throw out the following:
1) S16-09 13.5.c.i specifically directs designers to use S136 for hot rolled sections with high slenderness components. Clearly, CISC thinks that the provisions for CFM and hot rolled steel are somewhat interchangeable. The suggested modification procedure is entirely manageable.
2) Way back when, I did some reading on the development of the ancient SSRC research that has informed our column design procedures ever since. I was shocked that all of the testing was done on very small sections (W8?). Meanwhile, folks are out there designing Mega-Columns for skyscrapers using those same "calibrations". I submit that, in all likelihood, the statistical variation within hot rolled steel design is probably as great as that between hot rolled and CFM.
3) I should make it clear that I would never propose using CFM procedures for hot rolled steel where analogous design procedures for hot rolled steel already exist. I'd only consider doing so if CFM code provisions/testing offered a solution to a design problem that AISC/CISC hadn't gotten around to addressing yet.
The greatest trick that bond stress ever pulled was convincing the world it didn't exist.
CELinOttawa
Structural
Not an unreasonable view there Kootk, so long as read with your final paragraph.
Note: Neglecting failure modes is more then statistical tuning. Human error is very real, and anything we can do to avoid unnecessary complexity is of value.
I like to implement KISS in my work, and adapting a code to cover another material's design is what I'd call a "fundamental design method error".
I do agree that where equivalent methods do not exist, it is not unreasonable to extrapolate design methods from another material's design. You're not going to find anywhere where S136 covers something that S16 does not, in my honest opinion.
Note: Neglecting failure modes is more then statistical tuning. Human error is very real, and anything we can do to avoid unnecessary complexity is of value.
I like to implement KISS in my work, and adapting a code to cover another material's design is what I'd call a "fundamental design method error".
I do agree that where equivalent methods do not exist, it is not unreasonable to extrapolate design methods from another material's design. You're not going to find anywhere where S136 covers something that S16 does not, in my honest opinion.
I agree with your last post across the board CEL. CFM codes have only really started to mature over the last decade or so. Hopefully there's nobody from University of Missouri-Rolla here to be offended by that statement.
The greatest trick that bond stress ever pulled was convincing the world it didn't exist.
The greatest trick that bond stress ever pulled was convincing the world it didn't exist.
- Status
Similar threads
- Question
- Question