Topic Thread

  • 1.  Ultimate Live Loads at 1.0

    Posted 07-21-2022 05:00 PM
    The road to LRFD is slow and arduous, but we're almost there with updated maps and 1.0 load factors for wind, snow (as of 7-22), and seismic.  I assume ultimate rain loads are around the corner.

    Now, dead load is an actual estimation of weight, so the 1.2 and 0.9 factors are important.  But live loads are straight out of the code, and would be the easiest to update.  Just multiply all of Table 4-1 by 1.6, or revisit those numbers if they're not giving a consistent risk of failure anyway.  They're probably not--and if we're embarrased to put the number 160 psf (equivalent to 30 inches of standing water) next to the word "corridors", well, maybe we should be.  (In case you can't tell, I'm fishing for a defensive Subcommittee member with these barbs: I know you're out there.)

    Load combinations in ASCE 7-22 are:
    1.4D
    1.2D + 1.6L + (0.5Lr or 0.3S or 0.5R)
    1.2D + (1.6Lr or 1.0S or 1.6R) + (L or 0.5W)
    1.2D + 1.0(W or Wt) + L + (0.5Lr or 0.3S or 0.5R)
    0.9D + 1.0(W or Wt)

    With ultimate-level rain and live load, combinations would be far more consistent:
    1.4D
    1.2D + 1.0L + 0.3(Lr or S or R)
    1.2D + 1.0(Lr or S or R) + (0.6L or 0.5W)
    1.2D + 1.0(W or Wt) + 0.6L + 0.3(Lr or S or R)
    0.9D + 1.0(W or Wt)

    Are there plans to update live loads to ultimate in ASCE 7-28?  Is there any inherent meaning to the existing floor loads such that the 1.6 factor has significance?  Eager to hear any and all thoughts on this topic.
    #ASCE7
    #ASCE7-22
    #LoadCombinations
    #LiveLoads

    ------------------------------
    Christian Parker EIT, P.E., A.M.ASCE
    Structural Project Engineer
    Washington DC
    ------------------------------


  • 2.  RE: Ultimate Live Loads at 1.0

    Posted 09-04-2022 10:07 AM
    Christian,

    As you are likely aware, the road to LRFD has been a long one, with some recent updates.  The road starts back in the 1960s, when ACI introduced "Strength" design as an appendix in 318-64.  In 71 this became the primary method and ASD moved to an appendix.  Steel was next, in the 1970s with some substantial help from the National Bureau of Standards (now NIST).  Movers and  shakers in this effort were Bruce Ellingwood, Allin Cornell, and Ted Galambos, but others participated as well.  NBS published a recommended strength basis for ANSI A58.1 (the precursor to ASCE 7) in June 1980 as SP577 - "Development of a Probability Based Load Criterion for American National Standard A58."  This seminal work is likely the best reference for the underlying basis of LRFD. It states the underlying  theory, the formulations of load and resistance factors, and the assumed distribution types and parameters for various loads as well as resistances.  The important thing to note is that the "nominal" load values (e.g. 100 psf Live Load in corridors) predated this and were the ones historically used.   The load factors established in the load combinations were selected to calibrate the designs achieved by LRFD and ASD, the assumption being that ASD presented an acceptable design in most cases, and the nominal values were ones the profession had grown comfortable with.  For the case of structural steel, the calibration was based on producing the same size beam under dead and live loading (of assumed magnitudes) as one would obtain from ASD.  Ted MK (Ravi) Ravindra and Ted Galambos published a paper on this in the structural journal in 1978.

    Initially the load factor on seismic loads was taken as 1.4, again calibrated against ASD seismic design of the era (which was based on the SEAOC/UBC provisions).  In 1998, ASCE 7 adopted a new seismic hazard model based on MCE shaking (similar to what is used now).  It was reasoned that the uncertainty in ground motion was so large that it was hard to justify a  specific load factor on seismic, so 1.0 was taken.  Using the time proven calibration, the values of the MCE and DE shaking parameters were selected to calibrate  with Zone 4 designs in the 1994 UBC.  In ASCE 7-10, the seismic hazard model was again adjusted so that rather than using 2,500 years as the mean return period for MCE shaking, the MCE return period was adjusted so that with a load factor of 1.0 on seismic, the risk of collapse would not exceed 1% in 50 years, given an assumed structural fragility.

    Also in ASCE 7-10, the  wind load maps were changed from 50 year MRI to the Risk Category based maps we have today.  It  was decided to use a load factor of 1.0 on wind, and  the MRI for the new maps was selected so as to produce the reliabilities indicated in ASCE 7 Section 1.3.  Note these are not really "ultimate" or "strength" loads.  They are just  the loading that produces  the desired target reliability.  Due to uncertainty in material strength, engineering abiltiy to compute wind loads and other factors, there is signifciant probability that a structure designed for these wind speeds would not fail.

    There present ASCE 7-22 committee expressed no interest in converting the Live loads to a "strength basis".  Factors include the "familiarity and  comfort" levels cited previously.  There has been discussion that the live loads specified in ASCE 7 may need updating for many occupancies given that the present loads were set to typical furnishings and equipment present in typical buildings many years ago.  There have been some proposals to undertake the  statistical studies necessary to do this, but no funding has  yet come forward.








    ------------------------------
    Ronald Hamburger P.E., F.SEI
    Principal
    Simpson Gumpertz & Heger
    Oakland CA
    ------------------------------



  • 3.  RE: Ultimate Live Loads at 1.0

    Posted 18 days ago
    Ron,

    I appreciate the thorough explanation, and I'm sorry for not responding sooner.  In fact, I wish this timeline was part of my undergrad curriculum.  Only after entering practice did I realize how little I knew about ASD, and how much my understanding of design loads and limit states suffered for it.  Thanks for filling in some of the remaining gaps in my knowledge.

    I opened this thread daydreaming about simplified load combinations, but I do think a detailed statistical study would be worthwhile for economy and consistency.  Floor systems make up the bulk of the structural material, cost, and embodied carbon of any multistory building so it's strange to me that there isn't more interest.  Consider a typical framing plan for a building designed to ASCE 7.  A steel building with 200 beams and girders on each floor might have 12 braces.  A 10,000 SF concrete floor plate might have 75 linear feet of shearwall.  With the extensive effort our profession has undertaken to improve the accuracy of lateral forces, I wonder why we haven't chosen to refine the loads that govern most of the structure.

    ------------------------------
    Christian Parker P.E., M.ASCE
    Structural Project Engineer
    Washington DC
    ------------------------------



  • 4.  RE: Ultimate Live Loads at 1.0

    Posted 09-05-2022 11:14 AM
    In my professional opinion, "uncertainty" is the key word when referring to the "MINIMUM" Design Loads for Buildings and Other Structures coupled with simplified analyses.
    ASCE states that "it does not intend, nor should anyone interpret, ASCE's standards to replace the sound judgment of a competent professional, having knowledge and experience in the appropriate fields(s) of practice … "
    With the exception of analyses associated with existing structures, I know of none (even when given all the facts) that would not eventually reach for and apply some factor of safety.

    ------------------------------
    James Williams P.E., M.ASCE
    Principal/Owner
    POA&M Structural Engineering, PLC
    Yorktown, VA
    ------------------------------



  • 5.  RE: Ultimate Live Loads at 1.0

    Posted 21 days ago
    James,

    Thanks for the response.  You make a good point, but for environmental hazards, the safety factor lives in the return period to establish uniform risk.  For example, I'm told that snow loads in ASCE 7-16 and earlier resulted in dramatically different probabilities of failure across different regions, because the 1.6S factor doesn't correlate uniformly to the risk.  Now ASCE 7-22 has completely redrawn the maps (with much higher ground snow loads) and removed the importance factor and the 1.6 factor from the load combinations, resulting in heavier roof designs than previously required in some areas and lighter designs in others.  Rather than being extraordinarily sure that a likely event won't cave in your roof, we are now fairly sure that an extraordinarily rare event won't.  Everywhere.

    I appreciate your comment about safety factors and uncertainty.  19th century bridge engineers had very limited understanding of stability limit states by modern standards, but the Pennsylvania Railroad specified FS=5 for everything.  The safety factor reflected the uncertainty in their design, and most of those bridges are still here despite all the section loss and unforeseen hazards.  The danger of fine-tuning the code too much is that we may convince ourselves we can see the future.  However, I tend to think that as long as engineers understand uncertainty and the intent of the code, it's reasonable to open doors for folks who want to sharpen their pencils.

    ------------------------------
    Christian Parker P.E., M.ASCE
    Structural Project Engineer
    Washington DC
    ------------------------------



  • 6.  RE: Ultimate Live Loads at 1.0

    Posted 09-06-2022 07:55 AM
    While live loads are defined, they are still unknown.  A heavy piece or furniture could be considered a live load, such as a piano.  The other consideration is impact, if I am in my basement and one of my kids jumps off the couch, it is like an earthquake overhead!  There is still a place for ASD when determining deflections and drift.  Considering 100 psf on corridors makes more sense when thinking about egress stairs... during an evacuation event, the loading can far exceed 100 psf.  If ASD vs LRFD causes confusion within the profession, we should consider how the public perceives our approaches and how we explain them.

    ------------------------------
    Chad Morrison P.E., F.ASCE
    Professional Engineer
    Greenville RI
    ------------------------------



  • 7.  RE: Ultimate Live Loads at 1.0

    Posted 21 days ago
    Chad,

    Thanks for the thoughtful response.  It's certainly true that we wrap a lot of big questions into our floor loads, and it may be true that the public cares more about minimizing risk from human-imposed loads than minimizing risk due to external hazards like earthquakes.  That's hard to quantify, but if we had statistically rigorous use data, we could target a higher reliability for floor loads.  We could also determine a serviceability/ASD reduction factor appropriate to the bell curve of actual events.  As far as the realism of corridor loads, it makes sense for the loads to be different, but you really need to stack people on top of each other to get close to even 40 psf.  Running may produce higher impact forces, but running people take up a lot more space too.  I don't know how it all balances out (maybe 100 psf is too low), but what's interesting is that no one seems to know.  As far as I can tell, our profession hasn't really studied the probability of exceedance for floor loads.

    We do see failures from overloading floors--to my knowledge not due to people but due to heavy storage.  For my part, I always try to consider a worst case scenario across the life of the building, but some possible scenarios are too extreme to consider.  I could design a single family house that is then converted to floor-to-ceiling storage of nuclear waste in glass barrels.  I might regret my choices, but I can't very well design every single-family house for Risk Category IV and 250 psf floor load.  It's an extreme example, but there's a lot of gray in between.  I mentioned the updated snow loads in my response to James; another interesting change is that different thermal coefficients can be assigned based on the level of insulation in the roof.  Would you design a roof joist such that a future tenant's energy retrofit changes your DCR from 0.95 to 1.10?  I wouldn't.

    I'm intrigued by that last sentence in your response, but I'm not sure I understand what you mean about public perception.  Can you elaborate?

    ------------------------------
    Christian Parker P.E., M.ASCE
    Structural Project Engineer
    Washington DC
    ------------------------------



  • 8.  RE: Ultimate Live Loads at 1.0

    Posted 17 days ago
    While I understand the pros and cons of ASD & LRFD, once AISC combined them into one manual, the differences began to disappear to all that remains is a conversion between the two.  What is the load and what is the capacity?  Well, that depends... the public (or client) demands a clear answer without an explanation of factors.

    How many inches of snow can the roof support?  How many people are allowed on the balcony?  Only after a collapse, does it become apparent... that's too much!!!

    So how do we reassure the public?  0.3 becomes a literal slush factor when talking about 30" of powdery snow vs. 10" of wet snow.  Are the codes changing for the sake of it or improving for more accuracy?  The goal should be transparency in our methods, not accuracy.  Accuracy takes a backseat to serviceability.  The most efficient floor design may be prone to vibration as it lacks mass that offer inherit dampening.  The perfectly designed roof may visibly deflect and remain within acceptable limits of code.  But are they acceptable limits for the public to feel safe?

    Can this arena host a basketball game during a New England blizzard?  That has been the same question, before and after 1978.  The answer should be yes, of course it is two, three, four, five times as strong as it needs to be!!!   While code is written for engineers to use, it still needs to be translated before presented to the public/client.  I think this leads to unnecessary confusion inside and outside the profession, but I may be oversimplifying.  Ideally, best practice should drive design, not code minimum.

    ------------------------------
    Chad Morrison P.E., F.ASCE
    Professional Engineer
    Greenville RI
    ------------------------------



  • 9.  RE: Ultimate Live Loads at 1.0

    Posted 12 days ago
    Ah, I see.  I tend to think of the ASCE 7 as a balancing act between ease of implementation and economy of design, but you make a great point about the importance of transparency.  I would argue that the current live loads are very easy for engineers to use (aside from those pesky load combinations), but not especially transparent.  Floor pressures don't evoke a firm image of acceptable uses to laypeople.  By comparison, it's easy to explain that a curtainwall connection is designed for XX MPH and a XXX-year storm.  The commentary often provides helpful background, but not so for Chapter 4--presumably because there is no consistent basis, per Ron's response above.

    I agree that serviceability isn't getting the limelight it deserves, and I'd like to see more framing intentionally designed in excess of strength requirements for vibration, durability, and redundancy.  Any reduction in live loads would result in more framing being serviceability-governed.  Some designs would slip through the cracks, but in general it would push engineers to design for the requirements needed in each specific case, as an alternative to using an inflated catch-all loading which insulates us from considering vibration.

    AASHTO LRFD uses an impact factor as well as a redundancy factor on live loads.  AASHTO live loads are also data-driven, although we should recognize that it's a lot easier to collect data on roadway usage than building usage.  Not a roadmap, but an interesting case study in a related sphere of practice.

    ------------------------------
    Christian Parker P.E., M.ASCE
    Structural Project Engineer
    Washington DC
    ------------------------------