Simplified Shear Design of Structural Concrete Members: Appendixes (2005)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

E-1 APPENDIX E: Field Performance Data and Practitioner Experience A survey of the design practice of 26 different state departments of transportation and federal lands bridge design agencies was conducted. This survey included both a written questionnaire and either a telephone briefing on the response to the questionnaire or a written response. The objective of the questionnaire was to determine the status of conversion to LRFD, identify specific problems and practices with respect to concrete element shear design, to ascertain preferences for shear design methodologies, and to provide a vehicle for organizations to express their opinion of the current LRFD shear design methodology. Of the 26 agencies polled, 21 responded, and these states are Alaska, Arkansas, California, Delaware, FHWA CFLHD, Florida, Georgia, Illinois, Kansas, Kentucky, Mississippi, Missouri, Montana, Nevada, New Hampshire, New Jersey, Oregon, Pennsylvania, Tennessee, Texas, and Washington. The questionnaire and the responses are included in Sections E.1 and E.2, respectively. A summary of the results of the questionnaire was presented in Section 2.4. E.1 Questionnaire for NCHRP 12-61 Simplified Shear Provisions Design: I. Does your agency currently use the AASHTO LRFD Specification, including the Modified Compression Field Theory (MCFT) shear provisions (Section 5.8.3.3)? II. Does your agency use any modifications to the provisions? (e.g. pre-set θ and β values, simplified approach to critical section determination, other) III. Have you had difficulty applying the LRFD MCFT shear provisions? IV. Does the lack of a unique solution for shear design cause problems for your agency? V. Do the LRFD MCFT shear provisions produce designs significantly different than the AASHTO Standard Specifications provisions for your bridges? VI. Does your agency use the 1979 Interim AASHTO provisions (or previous editions) for prestressed member shear design? (Allowed by footnote to Section 9.20 of the Standard Specifications) VII. Do you have a preference for approach to shear design? Vc+Vs+Vp; mechanically- understandable; relatively simple to apply; examples? VIII. What types of bridges are most often designed by your agency? IX. Have you encountered cases where a particular method of shear design eliminated a bridge type from consideration? X. What would you say is the most important design issue, with respect to concrete shear design, that your agency faces? XI. Do the LRFD shear design provisions increase the design time significantly? XII. Are the LRFD shear provisions acceptable and clear for continuous beam design?

E-2 XIII. Do you use the Segmental Guide Specification shear design procedures for any of your box girder designs? Field/Existing Structures: I. Has your agency had problems in the field that could be directly attributable to a given shear design approach? This can include any previously-used design procedure? II. Has your agency had problems in fabrication of precast elements or construction of any bridge elements that are attributable to a specific shear design procedure? III. Has your agency had cases where existing and/or recently-designed structures will not work or rate if load rated using the LRFD approach? E.2 Responses to the Questionnaire for NCHRP 12-61 Simplified Shear Provisions Design: I. Does your agency currently use the AASHTO LRFD Specification, including the Modified Compression Field Theory (MCFT) shear provisions (Section 5.8.3.3)? 1) Yes, all new bridges. 2) Yes, all new bridges; converting Bridge Design Manual to LRFD. 3) Yes, have changed to LRFD, for all but steel boxes. 4) Yes. 5) Yes, on approx 99 percent of new bridges; all but one designer has changed over. MCFT is nice because it can be used for many elements. 6) Yes. 7) Mostly converted to LRFD. 8) About 50 percent are done with LRFD; trying to figure out how to implement standard shear designs for I girders. 9) In transition, but have not done any new concrete designs with LRFD. 10) Some LRFD designs on newer structures; several LRFD designs with LFD check. Some strut and tie designs have been done, but that method has been found to be quite difficult. 11) In process of switching over to LRFD; have completed several designs to date. 12) Currently in transition to LRFD and in evaluation stage of whether to use MCFT or allow simplified method. 13) In process of converting, still fairly green with LRFD. 14) LRFD implementation is only in development stage; not adopted as yet for production. 15) Not at this time; have one bridge in design by LRFD right now. 16) No, doing first LRFD right now. 17) No; other programs have impeded implementation. 18) Have not converted yet; not enough support software. 19) No. 20) No. 21) Yes, since 1997.

E-3 II. Does your agency use any modifications to the provisions? (e.g. pre-set θ and β values, simplified approach to critical section determination, other) 1) No. 2) Yes, use simplification to avoid iterations for angle. 3) No, but some smoothing on β and values. 4) Yes. 5) No, use tables in document. 6) No. 7) No modifications to the shear provisions. However, do use software, and if it includes simplifications, then those are used. The simplifications that are in LRFD for non-prestressed members often are not of use, because the demand is higher; thus to keep members at a similar size to the Standard Specifications, complete advantage must be taken of all mechanisms on the resistance side. 8) Yes, Standard Specifications method for shear capacity may be used with LRFD loadings. 9) No, not there yet. 10) No in-house changes; not to the point where we would be comfortable making changes yet. 11) No, and in limited calculations done to date the crack angle was set at 45-degrees and the concrete contribution, β, set to 2. 12) No changes formulated, but would welcome any modifications or simplifications, including the ‘traditional’ method. 13) No, but we would certainly preset some of the values as soon as our experience shows how it can be done. One designer investigated the current simplified method, but it proved too conservative for a precast slab unit. 14) No. 15) No. 16) No. 17) Not yet. 18) No, but have used β = 2 and θ = 26.5 for non-prestressed cases. 19) No. 20) No. 21) No. III. Have you had difficulty applying the LRFD MCFT shear provisions? 1) Yes, the learning curve is steep, even with training. 2) Yes with non-prestressed elements, such as footings and pile caps. 3) Yes, not simple enough and not clear where the numbers came from. Nice that strut and tie is now included, but difficult to follow in practice. 4) Yes. 5) Many of the younger engineers, who learned MCFT in school have had no problems. They have had more problems with Vci and Vcw method. There have not been any problems with checking. 6) General problem with LRFD is that it is a ‘black box’.

E-4 7) Absolutely. Several days of training by M. Collins helped, but still hard to pick up. Can’t apply it by hand, thus hard to check. Code not clear on how to apply. Use of simplified method too conservative when coupled with increase in demand that is part of LRFD. Additionally, we have not been able to arrive at standard shear steel designs for P/S girders, which was easily accomplished with the Standard Specifications methods. 8) Yes, complex. Difficult to apply by hand. Not a lot of confidence with the method, because it is not tested in field. Why throw out all previous experience? Method takes more time. Still trying to arrive at standard shear designs for girders. Not sure that you get a better design with LRFD methods. Software not there yet. 9) Don’t know yet. 10) Strut and tie difficult to apply, examples are always simply supported. Have had problems in continuous members. 11) Not applicable. 12) The short answer is a qualified yes. It takes some serious academic study to understand the new shear provisions. We are quite certain the provisions are a more accurate description of true behavior. However, as written, they are not appropriate for a design code. We opine that codes which work for the everyday designer are short, to the point, and also incorporate new knowledge all at the same time. 13) Several designers have indicated that it is too busy of a process for most ‘bread/butter’ type of structures. It requires software to do the iterations and can’t readily be hand checked. 14) Yes, as you are aware, Dr. Collins has produced several generations of the equations in LRFD. All are a pain. 15) We have had trouble applying all the LRFD Specifications. Have had trouble developing in-house software due to complexity. 16) No. For new designers it takes about the same effort to learn as Vci and Vcw method. However, some quick and intuitive checks would be nice. 17) Not applicable. 18) Iterative process difficult, especially without software. 19) No, we have not implemented LRFD design yet. 20) Not applicable. 21) No, however it is not the easiest thing to understand the first time using it for design. IV. Does the lack of a unique solution for shear design cause problems for your agency? 1) Yes. 2) No. 3) No. 4) Not enough experience to answer. 5) No. There have not been any checking problems and iteration is not an issue. 6) No. 7) No unique solution is a problem. 8) Have not experienced any problems yet.

E-5 9) Not there yet. 10) Yes for the strut and tie models when checking, otherwise not sure yet. 11) Not applicable. 12) Yes, code should be short and to the point. 13) To busy of a process, but otherwise not a problem. 14) Very much so. We need to be able to quickly spot check designs. When all possible paths have to be evaluated, this takes too much time and makes errors hard to spot. 15) No, not necessarily. 16) Not yet. 17) No. 18) Iteration adds effort and makes it more difficult. 19) It does have the potential to complicate the design. 20) Yes. 21) We are unaware of any problems. V. Do the LRFD MCFT shear provisions produce designs significantly different than the AASHTO Standard Specifications provisions for your bridges? 1) Not tracked. 2) Not getting very different results. Second edition changes made results similar to Standard Specifications. 3) Much better with new tables, closer to old AASHTO. Before steel at quarter point was a concern, now distributed more toward ends. Some congestion at ends, but not so much of a shear steel problem. 4) Not anymore. 5) In first edition there was a noticeable increase in shear steel, but now not a big deal. There is more steel near the quarter points. 6) No. 7) More shear steel, as much as 40-50% more with skewed bridges. Sometimes have had to add strand. 8) No. 9) No. 10) It requires significantly more steel than the Standard Specifications. This may occur for stirrups, for longitudinal steel, or a combination of both. Often see more stirrups within ‘d’ of the beam end. 11) Not applicable. 12) At this time, we are not certain about this question in regard to the resistance side of the equation. At this juncture, we have more concern about the significant increase in design shear forces on the loading side. 13) Not enough experience to answer yet. 14) ‘Significant’ is relative. It does cause additional stirrups in some areas of a beam. 15) Don’t know. 16) Not that we have noticed. 17) Probably not. We are conservative because of seismic. 18) 10-15% increase in stirrups at supports. 19) We have not investigated the design differences.

E-6 20) Not sure. 21) There have been minor differences in the amount of stirrups used, but nothing significant. VI. Does your agency use the 1979 Interim AASHTO provisions (or previous editions) for prestressed member shear design? (Allowed by footnote to Section 9.20 of the Standard Spec.) 1) No. 2) Yes, 1979 method is embedded in girder design program. 3) No. 4) No. 5) No. 6) No. 7) No. 8) Yes in past, but it has not been used in recent past. 9) No. 10) No. 11) Yes. 12) No. 13) Yes, exclusively, until conversion to LRFD. 14) Yes, exclusively and without any ill effect. 15) No. 16) No. 17) No. 18) Yes, occasionally, but P/S usage is small. 19) Yes, on simple spans only. 20) Yes. 21) No. Not familiar enough with these provisions to know if we used them prior to LRFD implementation. VII. Do you have a preference for approach to shear design? Vc+Vs+Vp; mechanically- understandable; relatively simple to apply; examples? 1) Simple is better, even though recognize that nearly all methods require software today. 2) Like the MCFT, because concrete contribution and tension are linked. Once familiar with MCFT it is easier. New engineers have trouble with Vci and Vcw method. 3) Need a simple method. The MCFT concept is OK, but how it has been put into the code needs better documentation and peer review. Changes in table values have raised concerns with engineers. The method for design should be easy to understand and give a feel for behavior, something similar to rectangular stress block for bending. 4) No.

E-7 5) The method should be easy to apply without mistakes. The Vc+Vs+Vp approach is a good way to present the method. Designers do not always care about the method being completely mechanically-understandable. 6) Method should be simple, not a black box. 7) Yes, simpler is better. 8) Iterative part is a problem in current method. 9) Simpler is better. 10) Prefer simpler methods, and iteration is distasteful. We are comfortable with the Vc+Vs+Vp concept. 11) Relatively simple Vc+Vs+Vp would be preferred. 12) We would welcome any simplifications. We are certain that the provisions are a more accurate description of true behavior. However, as written, they are not appropriate for a design code. 13) Should still allow simpler methods. If amounts of steel required between several shear design methods, why use the complicated method? You don’t need extreme accuracy. 14) Yes, but we give up a lot since concurrent web shear and moment cannot be obtained for continuous bridges. 15) The old AASHTO Specifications were easy to use. 16) Yes, for checking. With simpler methods can see contributions of different mechanisms. 17) All of the above. 18) Keep something simple in the specifications. 19) The 1979 and current standard specifications have worked well for us. 20) Yes. 21) The Vc+Vs+Vp approach seems straightforward enough, even with the complexity of calculating the individual terms. VIII. What types of bridges are most often designed by your agency? 1) P/S girders, boxes and slabs. Generally simple span designs; we are not thrilled with continuity due to detailing concerns. 2) P/S girders. 3) AASHTO Type III and VI girders, bulb tees, and precast segmental. Many are simple span made continuous, but not counted on. 4) P/S concrete. 5) Bulb tees, decked-bulb tees and WSDOT series P/S girders. Many simple spans, some continuous for live load. 6) Not a lot of concrete bridges, but some P/S girders and some boxes. 7) P/S girders 80-90%. Simple-span for both dead load and live load. Deck made continuous with crack control steel, but not counted in design. 8) P/S girders. Both simple spans and simple span made continuous for live load are used. Most simple spans have a continuous deck that is not counted. 9) About 50 percent of new bridges are concrete. Mostly precast ‘New England’ bulb tees. 10) P/S haunched slabs, P/S I-girders and K-sections. Just beginning to use PT slabs. 11) Prestressed concrete I-girders or steel plate girders are the majority.

E-8 12) Concrete deck on stringers, and concrete deck on prestressed I-beams and bulb T- beams. 13) Concrete CIP slabs, box girders, I shapes, bulb tees. 14) Simple span for dead load, continuous for live load prestressed beams. 15) Prestressed beam bridges. 16) CIP PT box girders and some P/S girders. 17) CIP PT box girders, slabs, and some P/S girders, which are made continuous for live load. 18) Steel composite with concrete deck, with some P/S girder bridges. 19) CIP deck with P/S concrete I girders continuous for live load. 20) P/S girders. 21) Concrete box beams (either adjacent or spread) for any span 25-feet or greater. Rigid frames, box culvert and the occasional timber bridge for shorter spans. IX. Have you encountered cases where a particular method of shear design eliminated a bridge type from consideration? 1) No. 2) Not yet. 3) No. LRFD shear design seems OK, particularly after modifications. 4) No. 5) No types eliminated. 6) Possible elimination could have happened, but not sure. 7) No, but for high skews shallower P/S beams may no longer work due to increased shear demand from new distribution factors. 8) Don’t know. 9) Not yet. 10) No. 11) No. 12) No. 13) No. 14) No. 15) No. 16) No. 17) No. 18) No. 19) No. 20) No. 21) No, but the small structures we typically do (<100 feet) the bridge type is usually set during scoping (and similar to previous projects), so shear design would not be a determining factor in the type of structure used. X. What would you say is the most important design issue, with respect to concrete shear design, that your agency faces? 1) The correct calculation of Vc, particularly after the section cracks.

E-9 2) Shear design for the negative moment region, and the longitudinal reinforcement design at mid-span. 3) Getting the calculations correct and being confident about the result. 4) How to apply the new specifications to existing bridges logically. 5) Load rating, particularly as related to the distribution of shear demand. 6) Understanding what appears to be a ‘black box’ method. 7) Trying to arrive at a standard shear design for all girders of a given size, which was possible with the Standard Specifications, but has not been so far with LRFD. 8) Need to have a simplified design method for checking. 9) A simpler method. 10) Developing a comfort level with the new methods, the learning curve. The lack of a hand method is an important issue for shear design, too. 11) Not applicable yet. 12) Larger specified design shear forces and complex shear resistance provisions. 13) Need a method to verify your design by hand. Too dependent on software, which may all come from one source. 14) The most important issues are speed, simplicity, and repeatability from one designer to another. 15) Simplicity, ease of use, and the ‘black box’ effect of the current method. 16) Keeping the designs easy to produce. 17) Congestion of shear reinforcement at bent caps, i.e. constructability. Also, not being able to check by hand is a concern, and software should show physically what is going on. 18) Simplicity and hand check methods with no iterations and one answer. 19) Not aware of any specific shear design issues. 20) LRFD needs to be simplified. 21) No critical issues that I am aware of. XI. Do the LRFD shear design provisions increase the design time significantly? 1) Not really, design has to be automated; so design time is then not an issue. 2) Not with a computer, and not any more than Vci and Vcw method. Hand checks do take more time though. 3) Yes, if you want to understand the result; no, if you just use the software. 4) No. 5) No increase in design time, difference is ‘peanuts’ with respect to previous method. In fact, many of the same parameters are required in both methods. 6) Don’t know yet. 7) Some increase, but not major. 8) Yes, anything in LRFD has increased design time. 9) Don’t know yet. 10) Design time is increased, even after the learning curve is accounted for. 11) They would increase design time, but once a process is laid out, I do not foresee the increased time to be significant. 12) Yes; however, time is not as important as clarity and understandability of the provisions themselves. 13) By hand, yes; by software, no.

E-10 14) Yes. 15) I think so. The LRFD Specifications increase the design time in almost all areas. 16) Maybe a little, but that is difficult to discern from the learning curve. In terms of total design, not that big a deal. 17) Not applicable. 18) Yes, especially without software. 19) It appears so, yes. 20) Probably. 21) No, we have beam design programmed into MathCAD. There was some initial programming time, but now the user need only fill in variables that change from project to project (θ and β, etc.) XII. Are the LRFD shear provisions acceptable and clear for continuous beam design? 1) Not sure, probably they are, but they have not been used a lot. 2) No, particularly in the negative moment region. 3) Strut and tie method is a good addition, but frustrating to follow, particularly for continuous beams. 4) Yes. 5) Provisions need to be clarified for continuity cases, in general. Change of calculated longitudinal strain from tension flange to mid-depth helped. 6) Don’t know yet. 7) Provisions are not clear. PCI examples are also not clear. 8) Don’t know yet. 9) Don’t know yet. 10) No, they are not clear to us. 11) Not applicable. 12) No. 13) Don’t know yet, have not done any continuous beams with LRFD. 14) No, the same issues of concurrent shear and moment still exist. 15) Don’t know. We have not done any continuous beams using LRFD. 16) Yes. 17) No, but better if proposed revisions go through in 2004. 18) Don’t know yet. 19) Future clarification would be helpful. 20) No. 21) The vast majority of bridges are single spans. To my knowledge, we have not done a continuous beam design within the last several years, since we have been using LRFD. XIII. Do you use the Segmental Guide Specifications shear design procedures for any of your box girder designs? 1) No. 2) Only on segmental bridges, and spliced-girder bridges done to LRFD. 3) No. 4) No.

E-11 5) No. 6) Yes, for one bridge. 7) No. 8) No. 9) No. 10) No, we very rarely, almost never design box girder bridges. 11) Not applicable. 12) No. 13) No, not in this office. 14) No, we do no segmental bridges. 15) We haven’t done any segmental box girders. If we do, we would use the Guide Specifications. 16) No. 17) No. 18) No, we have never designed a box girder. 19) No. 20) No. 21) Unfamiliar with this specification. Field/Existing Structures: I. Has your agency had problems in the field that could be directly attributable to a given shear design approach. This can include any previously-used design procedure. 1) Yes, but not any designed with LRFD. Some 1950’s and 1960’s vintage bridges have inadequate shear reinforcement that has led to a retrofit program. This was also partly due to heavier trucks and increased tire pressures. 2) No. 3) There have been problems with using the LRFD provisions for two segmental bridges, potentially stemming from the lack of principal tension stress checks in the webs. Additional PT (about 50 percent more) was added during construction of one of the bridges. 4) No. 5) One older bridge in inventory has had a problem with growing shear cracks. This apparently was due to significantly heavier truck weights than used in the design. The bridge has now been replaced. 6) No. 7) No. 8) No. 9) No. 10) Have had problems with cracking across hooks at bearing seats, which apparently is a development problem. This has occurred on 1930’s vintage simple-span T- beam bridges. Additionally, have had some diagonal cracking near the dead load inflection points of 1950’s vintage bridges, and this has required retrofit with epoxy shear bars. 11) Low shear ratings (using Virtis) have occurred in negative moment regions of structures designed with Load Factor Design. 12) No.

E-12 13) No. 14) None, including those designed prior to the drastic reduction in allowable concrete shear. 15) We have had some shear cracking in T-beams and bridge edge beams that were designed before 1976. That was about the time that the allowable shear was lowered to its current value. Since that time we have not had a problem. 16) No. 17) No. 18) Have had some shear cracking in negative moment region of continuous P/S girder bridges. 19) No. 20) No. 21) Not to my knowledge. II. Has your agency had problems in fabrication of precast elements or construction of any bridge elements that are attributable to a specific shear design procedure. 1) Not aware of any specific problems, but stirrup spacing is often tight. 2) Have had problems at the ends of P/S girders, where the splitting reinforcement (4 percent of P/S) is supposed to be located within h/5 of the end. Proposing h/4 to ease the congestion. 3) No. 4) No. 5) Congestion of shear steel usually not a problem. Congestion of confinement steel is a problem, however. 6) No. 7) Not yet, but LRFD method may produce unacceptable shear steel quantities for high skews and shallow beams, which worked by Standard Specifications. 8) Hard to say, because it is difficult to trace problem back to cause. 9) Have had diagonal cracks in webs near ends after strand release. Could be fabricator problem. 10) No, but have used self-leveling concrete to avoid compaction problems in congested areas. 11) Not applicable. 12) No. 13) No. 14) Not directly. Any prestressed beam with deflected strands requires dense patterns at the beam ends. 15) No not generally. There is a slight problem in getting all the shear steel into the ends of prestressed beams. It is getting quite congested. 16) No. 17) Joint shear designs of column footings or pile caps using some of the earlier seismic criteria were too congested. 18) No. 19) No. 20) No.

E-13 21) Not necessarily specific to a shear design procedure, but we have had some difficulty with the number of stirrups and the amount of end block steel conflicting at the ends of box beams. It has led to fabrication problems, either with changes at the shop drawing phase or adjustments made during fabrication (i.e. stirrups at spacings different from the plans or stirrups that shift in the form and create groups of bars). We have made an effort with recent box beam project to reduce the amount of overlapping or repetitive end block steel so that there is more flexibility in its placement and so that the stirrups are placed correctly. III. Has your agency had cases where existing and/or recently-designed structures will not work or rate if load rated using the LRFD approach? 1) Not using LRFD for load rating (LRFR), yet. 2) LRFR will be a problem if newer bridges don’t ‘rate’. AASHTO guide is forthcoming. 3) There is a real debate over LRFR right now, may use older specification for rating older bridges. 4) Yes, almost every multi-column bent. 5) Some load ratings must use grid models to develop actual shear force distributions. 6) If load rating is a problem with new specification, then go back to older specification that worked. 7) This could be a problem, although load rating is done solely on flexure, here. 8) Don’t know yet. 9) Not using LRFR yet, but this likely will be an issue. 10) We are just starting to explore LRFR now; thus, no problems yet. 11) We have not rated structures LRFD, as of yet. 12) Not sure. 13) Don’t know yet. 14) We have not made specific checks. Others have reported such problems exist. 15) There is lots of controversy with LRFR. LRFR is not ready yet. 16) Have not looked into this yet. Whether it’s a problem depends on how the system is set up. 17) Not applicable. 18) This would be a problem if LRFR caused something not to rate that previously did rate. 19) We have not investigated existing structures using LRFD. 20) Don’t know. 21) Our ratings continue to be done in LFD. There was one LRFD designed bridge that had a perfectly acceptable beam design using the HL93 truck and the design tandem. However, when rated using LFD and one of the Delaware design trucks (S335 – a tandem load), it rated at just under 1.0. An additional pair of strands was added so that the design met both criteria.