Cast-in-Place Concrete Connections for Precast Deck Systems (2011)

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.

C-i Appendix C Phone Survey Results

C-1 C.1 Introduction to Phone Survey This appendix provides a synthesis of the comments received from the individual respondents who participated in the phone survey that was used to gather information including the respondents’ experience with similar systems, their input on important performance criteria, and their feedback on proposed connection concepts. Nearly 60 people were interviewed during the phone surveys which were conducted in partial fulfillment of project Tasks 1 and 2. The individuals interviewed were selected through consultation with the project team, David Beal, and interviewees who recommended others who should be contacted. The respondents represent bridge engineers (including many individuals who serve as State Bridge Engineers), consulting engineers, fabricators, material suppliers, industry representatives, and technical committee contacts. The distribution is presented in Table C.1.1. Table C.1.1: Distribution of phone survey respondents B rid ge E ng in ee r M at er ia l S pe ci al is t Fa br ic at or R es ea rc he r In du st ry R ep C on tr ac to r O th er 36 9 6 13 3 0 3 A few survey respondents were identified under multiple categories in Table C.1.1, so the total number in the table (70) exceeds the number of survey respondents (59). Also note, that although there are no individuals listed under the “contractor” category, a number of bridge engineers and fabricators were able to provide firsthand experience on constructability issues encountered in the field. Attempts were made to include contractors in the survey, but because of the high level of activity with the contractors in the field towards the end of the construction season, and the difficulty in reaching the individuals in the field, this was not possible. The contractor on site during construction of the MnDOT implementation of the Poutre Dalle (PD) concept was queried by members of the project team regarding constructability issues, which influenced some of the decisions made in the concept development implemented in the University of Minnesota laboratory specimens described in Chapters 5 and 6. A complete list of the phone survey participants is included in Table C.1.2. Prior to being interviewed, the participants who agreed to be surveyed were sent a packet that is included at the end of this appendix. The package included brief background information, figures of proposed concepts, and sample questions to be posed during the call. Participants were told that information hoped to be gained from the survey included: • Details of similar precast bridge deck systems in use or under consideration, including experience with the systems and issues encountered during design, fabrication, construction and service. • Input on key performance criteria to be used to evaluate proposed concepts, and • Feedback on concepts considered by the research team for investigation as part of NCHRP 10-71.

C-2 Table C.1.2: List of Phone Survey Participants First Name Last Name Br idg e E ng ine er M at er ial Sp ec ial ist Fa br ica to r Re se arc he r Ind us try Re p Co nt ra cto r Ot he r Employer Sameh Badie X George Washington U Shri Bhide X X Portland Cement Association Gregg Blaszak X X Tech Fab Hank Bonstedt X PCA Pennsylvania Jimmy Camp X P.E., New Mexico DOT Reid Castrodale X X Carolina Stalite Co. Vijay Chandra X Parsons Brinckerhoff Ned Cleland X Blue Ridge Design, Inc. Mike Culmo X CME Associates John Dick X Director, Transportation Systems PCI Dan Dorgan X MN DOT Marc Eberhard X U of Washington Al Eriksson X Splice Sleeve North America Merv Eriksson X USFS Roy Eriksson X Eriksson Technologies, Inc. Sam Fallaha X Nebraska Dept. of Roads James Flynn X NYDOT Ajit Gokhale X FLUOR-OBDP - Salem, OR Jon Grafton X X Pomeroy (producers) Robert Gulyas X Master Builders Joey Hartmann X X FHWA Neil Hawkins X Prof. Emeritus University of Il l inois Mark Henderson X LJB Engineers and Architects Finn Hubbard X Wisconsin DOT Michael Hyzak X TX DOT Mohsen Issa X University of Il l inois - Chicago Steve Johnson X South Dakota DOT Keith Kaufman X Morse Brothers, Inc. Bijan Khalegi X WS DOT Tom Koch X NC DOT Mervyn Kowalsky X NC State Univ. Dept of Civil Engineering Jim Ma X CalTrans Kevin MacDonald X CEMSTONE Lee Marsh X Berger/ABAM Engineers Elmer Marx X Alaska DOT Norm McDonald X Iowa DOT Don Meinheit X X Wiss Janney Elstner Associates Rich Miller X University of Cincinnati Will iam Nickas X FL DOT Mike Oliva X University of Wisconsin Vic Perry X DUCTAL Charles Powell X ThermalChem Chuck Prussack X Central Pre-Mix Prestress Basile Rabbat X Portland Cement Association MaryLou Ralls X former TX DOT Russell Robertson X FHWA Carin Roberts-Wollmann X Virginia Tech Dept. of Chem. Eng. Hugh Ronald X HDR Engineered Structures Henry Russell X X X Henry Russel Inc. Andrea Schokker X Penn State University David Scott X NH DOT Steve Seguirant X X Concrete Technology Corp. Larbi Sennour X Consulting Engineers Group Peter Smith X X Fort Miller Company, Inc. Maher Tadros X X U of Nebraska, Lincoln Benjamin Tang X FHWA (NY) Julius Volgyi X VA DOT RIch Zeldenrust X Washington State DOT Steve Zendegui X Jacobs Civil , Tampa TOTAL 36 9 6 13 3 3

C-3 C.2 Performance Requirements/Specifications The respondents were asked to comment on design/fabrication/construction/performance criteria that must be satisfied to provide adequate performance of key parameters to be included for consideration on this project (listed below). The parameters listed in parentheses are sample issues that must be addressed. • Durability (Cracking; Corrosion resistance) • Strength (Static strength; Continuity for live load; Load distribution) • Fatigue resistance • Seismic resistance (Strength; Ductility) • Constructability (Transportability including minimization of thin sections and size of precast concrete pieces; Ease of installation including accommodation of differential camber; Adaptability including skew or transverse slopes and ability to accommodate future expansion) • Rapid construction (Speed of erection; Reduction of formwork; Field placement of reinforcement) • Serviceability/Performance (Rideability) • Economy (Construction, Life-cycle operations/maintenance) Respondents were first asked to indicate whether or not they thought anything was missing from list. They were then asked to rank, in order of importance for the project, the key parameters identified above. Regarding whether or not anything was missing from the list, most respondents thought that all of the key parameters were identified. A few of the respondents listed additional issues of concern. These are summarized in the following section. C.2.1 Additional issues identified for consideration Associated with durability, staining was listed as an issue of concern that is a result of cracking and subsequent leakage through connection regions. Regarding the issue of fatigue, M. Eriksson [USFS] indicated that for the systems to be considered on low volume roads, fatigue should not be considered as a criterion of concern. Under the category of constructability, it was suggested to consider two components: (1) access to site (e.g., in remote areas, bulb-Ts may be preferred because they are durable and robust in transport, and accessibility to CIP on site may be impossible or outrageously expensive); and (2) reducing construction time in field, which though broadly applicable, has dramatic implications at remote locations or regions with short construction seasons. It was suggested that consideration should be given to developing systems that put more time into fabrication at the plant to reduce the fabrication time required in the field. Under the category of economy, many respondents indicated that everyone talks about the importance of life cycle costs, but in reality, initial cost is what is important to state DOTs, because of limited

C-4 budgets and numerous projects that must be funded. Culmo indicated that it should be noted that rapid construction automatically adds 15-25 percent increase in cost. Matching the joint color with the precast so the systems do not appear jointed was suggested as an added criterion by Bhide and Ralls. Culmo [CT, NH] also suggested including inspectability and tolerances (including sweep and camber) in the list of performance criteria. He felt that the tolerances issue needed to be a major bullet item and should include vertical and horizontal adjustability that must be built into the system. Tolerances In addition to the comments by Culmo, several other respondents commented on tolerances. Dick (PCI) indicated that tolerances, alignment and fragility all needed to be evaluated. Flynn (NY) indicated that tight tolerances are needed. Most states have specific girder tolerances and total out-to-out tolerances. Tolerances include sweep, width, depth, flange thickness, etc. (e.g., specifications can be found on AK website). Tolerances in PA were initially based on PCI product tolerances; however, when the state experiences a new problem with a product, they will generate a new tolerance requirement (e.g., plumbness of web). Some states (e.g., NM) have had problems with tolerances in terms of camber and sweep [Camp]. Tang (NYDOT) indicated that sections often have a lot of sweep which makes forming joints in the field difficult. Ralls indicated that it would be desirable to have more tolerances in the field (e.g., FL built a 143 ft. span that had to be placed with just 1 in. to spare on each side). Roberts-Wollman indicated that issues associated with tolerances are exacerbated if the systems do not incorporate overlays. In addition to consideration as a performance criterion, Smith stated that tolerances are needed in the specification. He believed the loop bar detail would be good for longitudinal joints because he thought it could readily accommodate tolerance issues. He also indicated that wider joints (e.g., ≥12 in.) can make up the difference in camber among adjacent precast elements. C.2.2 Rank ordering Several of the respondents were asked to rank order the performance criteria in order of importance (with “1” being highest), assuming that the proposed connection concepts satisfied at least the minimum acceptable criteria in each category. In other words, all things being equal, respondents were asked whether they would prefer a system that provided a slight edge on durability or a slight edge on speed of construction, for example. Table C.2.1 contains the number of responses for each ranking value for each of the performance criterion. As noted below the table, if the respondents listed multiple performance criteria as equally important, those criteria were given the same ranking (e.g., if durability, strength, and seismic were selected as top choice, they would all be designated with a “1,” the next choice that would follow would be designated “4;” there would not be a listing with a “2” or “3” for that individual). Of the 59 interviews, 15 did not rank order the criteria. Of the 44 individuals who did offer ranking, it is clear that

C-5 durability was considered of greatest importance, as it was mentioned by 40 of the 44 individuals who ranked the criteria, and of those, 75 percent of them ranked it as the top priority. Besides durability, other issues considered of great importance were constructability and rapid construction; followed by economy and serviceability/performance. In reviewing some of the comments of the survey respondents, they felt that if the sections were not durable, none of the other criteria mattered. That is, although it is desirable to have bridges that can be constructed rapidly, if the bridges do not last, it creates more problems in the long run than if they had taken longer to construct a more durable structure in the first place. Strength was listed on the lower end of the spectrum, because many respondents indicated that they could always design the structure to ensure that it had adequate strength or find a means to achieve the required strength. So it was not that strength was not necessarily a primary concern, it was felt that it was not an issue. The lowest ranked criteria were fatigue and seismic resistance, in fact for these categories, fatigue and seismic were listed by 6 and 11 individuals, respectively, as either low or not applicable. In the case of fatigue, one of the individuals who ranked fatigue as a high priority (i.e., rating = “2”) was concerned with the potential fatigue resistance of the WWR connection detail in particular. Some of the respondents amplified their responses with additional comments on their selections. The following selected comments provide insight into some of the rank ordering. This is followed by some grouping of comments associated with particular topics (e.g., seismic, rapid construction, etc.) in cases where a number of people provided input on similar issues.

C-6 Table C.2.1: Rank ordering of performance criteria Rank Order Durability Strength Fatigue Seismic Constructability Rapid Construction Serviceability/ Performance Economy 1-highest 30 3 1 0 13 9 8 6 2 4 3 1 0 9 8 2 4 3 4 5 1 0 7 3 2 3 4 1 3 2 2 0 0 5 3 5 1 0 2 3 1 3 0 2 6 0 1 0 2 0 2 1 1 7 0 0 1 2 0 0 1 1 8-lowest 0 0 1 0 0 0 0 0 SUM 40 15 9 9 30 25 19 20 % of respondents 91 34 20 20 68 57 43 45 1If the respondents listed multiple performance criteria as equally important, those criteria were given the same ranking (e.g., if durability and strength were selected as top choice, they would both be designated with a “1,” the next choice would be designated “3”…there would not be a listing with a “2” for that example).

C-7 General comments associated with performance criteria The following is a list of comments which amplify the respondents ranking. The comments are ordered alphabetically by State DOTs, followed by alphabetical listing of other respondents. AK indicated that as part of the durability criterion, crack width (see AASHTO serviceability), coulomb resistance to chlorides, elasticity of grouts, greater ductility and resistance to abrasion are important. Rapid construction is important to Alaska because of the short construction season (March-October). CA believes durability (including cracking and corrosion resistance) is an issue of particular importance for joints. CATrans has closure pours when widening bridges and cracking and corrosion that can occur at those locations is of concern. Rob Chai (UCDavis) is currently conducting research related to these issues. They have discussed using stainless steel, FRP and configuring steel layout to achieve minimum width of closure pour. IA DOT considers that if economy is not there, it is hard to justify any of the proposed systems. IA does not have a concern for rapid construction [McDonald]. Dorgan (MnDOT) indicated durability and low or minimal long term maintenance go hand-in-hand. With limited budgets, first costs have a huge impact on the choice of system. For rapid construction definition, on local roads, it means cutting construction time in half. For interstate highway bridges, rapid construction could mean a weekend. Although fatigue may not be a big concern, it should be considered with regard to breaking up grouted keyways and effects of differential deflection on further breakup. NYDOT considers constructability and durability as top priorities. They use stainless steel where they require better corrosion resistance and longer life. They also use corrosion inhibitors with HPC in precast elements for good long term performance, and seal decks with a silane sealer. PA does not do continuity for live load. Criteria such as strength/fatigue/seismic are engineering requirements that must be met by all systems (they are overarching, rather than fitting them into rank order scheme of priorities) [Bonstedt]. Rideability is a criterion that can be fixed (i.e., you can overlay or grind it). TX is not in extreme corrosive environment except the panhandle. The important performance criteria for TX include constructability and transportability [Hyzak]. Keep widths of sections down for transportability. Durability is also important for the joints. They will be the “weak link.” Ralls mentioned staining and rideability as important issues. She indicating cracking is a problem, but the real issue is corrosion. Rather than using corrosion resistant materials, she would rather have the focus on high performance concrete. For rapid construction, Ralls indicated the less concrete the better. For serviceability, a wider joint width helps rideability and grinding can be used. When all is said and done, economy (initial cost) is still the driver. Strength and lateral load transfer can be achieved. Robertson (UT) indicated that constructability issues to be considered included traffic, availability of CIP, room to build (i.e., right-of-way). Volgyi (VADOT) indicated the top priority would be cost (like it or not first cost controls). Durability is also a top priority. He has concerns whether or not the systems can hold up without PT.

C-8 Zeldenrust (WSDOT) listed performance/durability high on the list. Cheaper and faster are not always best. Regarding durability, the concern is whether or not the system is leakproof. It must behave well in the long term and at least as good as monolithic—that should be the standard. Shipping and transportation is important. Initially fabrication will cost more in development of new form, but afterward, costs should level out. Bhide indicated that the priorities for the project research should be different than the priorities of bridge engineers which are (1-safety; 2-economy; 3-aesthetics). Castrodale indicated that issues related to serviceability include differential camber and camber growth. He also indicated that rapid construction is sometimes considerably slowed by requiring screeding, it is better to fill the joint and subsequently grind if necessary to achieve rapid construction. It is important to be aware of capacity of equipment needed (due to weight and size of pieces); availability of CIP due to distance from plant. Chandra indicated that durability is a concern when CIP is used. Type A cement is expensive and sometimes hard to find. Eriksson stated that it is a given connections must have durability/strength/serviceability/performance. Consider bridges for life (i.e., 75 year lifespan is desired); consider making quantitative crack width requirements; use better materials (i.e., good detailing and better grouts); and consider rethinking how live load is distributed within the system. Hartmann indicated all systems developed as part of this program must meet the objectives of rapid construction and economy. If they do not meet these objectives, it does not matter if they meet the second objectives (durability/constructability) because they would not be competitive. Hawkins indicated that constructability also involves adaptability (e.g., can the proposed system accommodate skews). Oliva indicated that flexural strength is not a key issue at joints, because most decks fail in punching shear. Other aspects to be included in criteria include surface roughness on girders and bond that can be developed. Prussack emphasized the importance of durability. If the bridge is put into service, and needs replacement a couple of years later, the speed of construction was not worth it. It is difficult to achieve, good, fast, and economical simultaneously. In considering speed of construction, the amount of time it takes grout to cure must be considered. It is important to know the objectives of the project. One-day installation might require magnesium phosphate grout. In order to move to a new system, the entire mindset must be changed. Rabbat thought that “leakage” should be considered under category of durability. Need to pay attention to type of grout chosen. Grout needs shrinkage resistance and workability (which are often competing parameters). It is extremely hard, if not impossible, to avoid cracking. Roberts-Wollman suggests figuring out what is causing cracking and find a way to eliminate corrosion. Freeze-thaw is probably not a top priority. She also indicated that serviceability and performance should include rideability.

C-9 The system has to be easy to use and result in reduction in labor if trying to compete with CIP [Ronald]. Use these criteria to screen out potential systems. If the potential systems satisfy these, then make them work to satisfy the other criteria (e.g., serviceability, strength, durability). Also consider climate as a criterion. If one system requires good weather and another is not as restrictive, the system which can be built in any environment is the preferred option. Russell commented that in Miller’s research on boxes, his conclusion was that temperature effects on the joints was significant. Precast systems should be designed to be placed at the site without concern for particular orientation—the installation should be fool-proof. If a section can be put in wrong, it is likely that it will occur in the field at some point. The majority of the work should be done in the plant off site such that limited work is required in the field. Initial cost is also important because of DOT limited annual budgets. Scott (NH) stated that all of the criteria are essential in design; fatigue and strength cannot be compromised. Seguirant indicated that durability is the primary concern with the systems. However, it is less of a concern with the PD system because of the thickness of the CIP over the joint which is likely to mitigate reflective cracking. Durability will likely be improved if joints are enlarged and filled with concrete instead of grout. Seigurant also indicated issues such as load distribution need to be addressed for some of the systems. AASHTO provides a simple calculation for decked bulb-Ts, which may not be very accurate. Regarding other criteria, constructability is a big issue. Transportability will vary from state to state because of equipment availability and permit requirements. Differential camber can be dealt with using leveling techniques. Decked bulb-Ts must be adaptable to different bridge geometries because the precast is the final bridge grade. Decked bulb-Ts have been manufactured to varying vertical curvatures by varying the thickness of the flange (although this adds weight and has practical limitations). Concrete Tech Corp has deflected forms to change profile of girder to achieve up to 2 ft. of vertical curvature. Skews and cross slopes are generally not a problem, although combined vertical curvatures and skew can lead to flange match-up problems. Smith indicated that under the criteria of durability, both cracking and corrosion resistance are important. The slab strength should be carefully analyzed as this has not been done in the case of the two bridges with loop bar details that Smith described; neither deck was analyzed for bending strength which may be an issue for the larger girder spacing. Regarding constructability, Smith did not see a need to get by without a form system. For the loop bar detail shown in the survey, he was more concerned with spalling chunks of concrete from the thin sections that butt; construction ease is not worth the risk. Smith also indicated that it is not wise to have a large percentage of deck area done with grout; it is difficult to do it right and achieve good durability. Smith does not like about transverse joints. Tadros stated that transverse jointed systems should be designed to provide adequate live load distribution. Without PT, he is concerned that the system will have transverse cracking. Reinforcement is required for crack control and corrosion protection. Poutre Dalle system may need to address fatigue for the joint over the pier. Joints should not be weaker than the rest of the system. Reinforcement should be placed as close to the top surface as allowed for crack control.

C-10 With regard to durability, Tang (NYDOT) suggests a “nested system” approach (i.e., epoxy-coated bar or stainless steel with HPC). Avoid forming in the field—make pieces that butt. Rapid Respondents had many comments on rapid construction. Rapid meant different things to different people. Simply shortening the typical time span would be desirable (e.g., instead of constructing bridge over one construction season, cut down time by half). Tadros believes that the primary performance criteria to be addressed is rapid construction/speed of erection. He also noted that when CIP is required, placing, finishing, curing take time—it should be avoided. He indicated deck bulb-Ts have a good future for rapid construction. He recommended keeping things simple. Ronald indicated that the concern should be on cutting the schedule of a conventional job. If it is an emergency repair, it is not unusual to pay a premium for such a system. Rather than focusing on niche markets (e.g., remote areas where CIP is expensive) develop systems that are competitive in regular market. Issues that speed erection include reduction or elimination of formwork, CIP, or onsite bar placement. C.3 Comments on Precast Slab Span Superstructure Concepts (e.g. Poutre Dalle System) The survey included renderings of the Poutre-Dalle concept and the MnDOT implementation of a Poutre Dalle concept as examples of precast slab-span superstructure concepts. Many of the respondents liked the Poutre Dalle concept, particularly the MnDOT implementation of it. Many mentioned potential applications for the system especially on low-volume roads. A few individuals recommended considering the use of voids or lightweight aggregate concrete (LWAC) to achieve longer spans with the PD system; however, a number of other individuals were strongly opposed to voids and LWAC due to problems encountered in the past. An issue that was noted several times in relation to the PD system is that the use of CIP makes this system less rapidly constructable. It was not necessarily the quantity of the CIP, but rather that it was used with the system at all, that they found would slow the construction time due to the need for curing and strength development. The following are some of the comments in alphabetical order in terms of the state agencies represented, followed by alphabetical listing of others who provided comment. Marx (AKDOT) liked the shallow span-to-depth ratio of the PD system. He thought it provides economy by requiring less fill in overpasses. In AK, however, CIP should be avoided because of its high cost in remote areas. Ma (CalTrans) was very interested in composite slab span systems. Standard details would be helpful to them for this type of system. He thought the PD looked attractive with the bottom flanges abutting. For short spans, he indicated the differential camber of the PD system should be small and there may be no need to debond strands at the ends. He did not see how the short clearance of the bars protruding from the web could provide horizontal shear transfer—it was not clear how they could be developed. Nickas (FL) had seen longer versions of the PD bridge. He thought the vertical upright bars in the MnDOT implementation would pose a safety issue. Inspectors in FL request bending all vertical bars down. He was also concerned with potential interference with the protruding bars. He thought that it could have

C-11 potential problems orienting the pieces in the field. In the range of 50-60 ft., FL is using the NE inverted Ts. Nickas would be interested in using the PD system in spans of less than 35 ft. He does not see a need to add wire mesh in the tips of the PD; bulb-T systems have larger areas without rebar than the PD detail and Nickas does not have memory of losing a chunk of concrete after construction was completed. McDonald (IA) thought that the PD system looked better than current inverted-T system they are using. For longer spans he recommended trying to reduce dead weight by considering voids. In addition, he recommended considering changing some reinforcement to make the precast section more constructable (e.g., change S1608E bars to two C-shaped bars). Kaufman (Morse Brothers, Inc., OH) had concern with PD handling skew and staged construction. Flange could be decreased down to 3 in. He did not think that making the flanges thinner would help with reflective cracking, to avoid reflective cracking he recommended putting a lot of reinforcement across the joint. Dorgan of MN is a proponent of the system. He favored working to further reduce mild reinforcement which would ease congestion. He preferred a roughened top surface of bottom flange as the system in the field currently appeared to be performing well. Falhalla of NE liked the uniform slab which is good for traffic. Scott (NH) was concerned whether precasters would buy into building new forms for the PD system. He was also concerned with so much CIP that the system would not be rapid. He wondered what the advantage of this system was over a butted box beam. Camp of NM saw a real advantage to the PD. He thought that it looked attractive and could be competitive with a large inventory of slab bridges NM builds in the range of 40-45 ft. Advantages include the need to pour less concrete and the elimination of formwork. He thought the change to 90 degree bends in the bars lapping in the adjacent sections was a great way to ease construction in the field compared with the original 180 degree bends used in the French implementation. Flynn (NY) did not see much of an advantage with the PD system. Complications of the system include the single protruding bar. He indicated that mechanical connectors should be explored; however, he preferred not to “spin” the bars on in the field. Koch (NC) thought the PD system would be a harder sell than most common precast systems because it appears to be more “exotic.” NC likes to try less exotic systems, however he thought the PD system looked innovative. He favored the shallowness of the section. They have used voided slabs on short span structures (e.g., 40-60 ft.) with a 2 in. asphalt or concrete overlay; most of these systems have performed well. Of 100 bridges reviewed, the ones with poor drainage issues (e.g., without grade or cross slope) had problems where water either sat or filtered down. Koch suggested LWAC might be used to lighten up the section. Bonstedt of PA liked the PD system, but thought it looked heavy. He thought the bars protruding from the form may be difficult to accommodate in the plant, but installation in the field did not seem difficult. He suggested laying out the reinforcement to avoid bars interfering in the field.

C-12 The PD system looked like much more work in field compared to double-Ts with weld plates that SD is currently using [Johnson]. PD looked good for its load transfer. If it could be used on state highways, it would be a benefit. Advantages include the elimination of formwork and good riding surface. They appear more expensive than their typical bridges. If the system is heavy, it would require more support at the abutment and piers than their conventional systems. Hyzak (TX) thought application of the PD system was limited because of short span lengths. Competition is prestressed slab systems. Likely applications for PD systems include stream crossings and locations where shallow clearances are required. Ralls thought the PD system could be improved if the pieces could be set straight down without sliding them under adjacent pieces. TXDOT has done some similar new shapes. The PD system is also more conventional than rapid, but allows for rapidity of construction through precast while including more conventional CIP. Volgyi (VA) stated that if rapid is the desire, then avoid use of CIP. He did not think that the PD system went far enough toward rapid construction with the CIP component. Zeldenrust thought the PD system had its applications, but it is limited in span length. A standard element used by WSDOT is a 2 ft.-2 in. voided slab. Khalegi of WS indicated MnDOT modified original PD in logical ways, except he did not favor the protruding bars. To increase potential span length, WSDOT proposes a voided system without protruding bars. WSDOT uses a solid slab system up to 50 ft. on local roads. WI also sees PD applicability for short spans, and would likely use it on local roads. Hubbard indicated lightening up the section while maintaining strength would be a good goal. Eriksson (USFS) thought PD looked like interesting concept but appeared heavy. He thought it might be difficult to compete economically. Also in remote areas CIP is not readily available. Bhide agreed the PD system looked good but heavy. In addition to considering voids, he recommended reducing the web width or using LWAC. Bhide also commented on the voided slab system developed by Tadros at UNE. Castrodale indicated that CIP of PD system hurts its efficiency; must wait 28 days for it to cure. Chandra was concerned with connection between the bottom flanges. Culmo liked the concept. He had some concerns with reflective cracking and the time required in the field for reinforcement and concrete placement. He recommended making the web full depth and pouring the joint above the flange in the field to eliminate the need for screed rails. A float finish should be all that is required. A watertight membrane and overlay on top would help deal with differential camber issues. Culmo believed this double field system would be faster than a single cast. He did not recommend the system for highway bridges. Dick was concerned with the weight of the system and thought that the proposed section was unnecessarily wide. He recommended modifying the bottom flange to be a SIP form rather than keeping the web as part of the system. Ericson thought the system was a perfect example of an emulative design.

C-13 Eriksson stated that the section seemed highly erectable and efficient; the drawback was the large amount of CIP. Gokhale, who participated in the PD showcase, did not see any advantage of this system to voided slabs. He thought voided slab construction would be quicker and there would usually be an asphalt overlay with a membrane or thick HPC deck on the top which he thought would require less effort than pouring the CIP portion of the PD. Grafton (Pomeroy) indicated the only change he would make to the PD system would be to eliminate the projection of the bar beyond the edge of the flange. The current configuration requires “drifting in” the precast pieces laterally in the field. He also suggested a flat seal or T-seal over the joint between the abutting edges. This system would be in direct competition with other short span bridges (e.g., voided slabs) in the range of 30-50 ft. Hartmann thought the MnDOT “drop-in” cage was a great improvement over the original concept. The section looked appropriate for the span range. Henderson and Sletten thought that the PD system would be very “doable.” They indicated that it was similar to box beam systems with a CIP deck on top. They thought the system would perform well, but may develop some reflective cracking. Issa thought the system looked good, but would not be rapid because of the CIP. He thought all of the suggested variations of the PD, described in the survey handout, would work. Marsh liked the MnDOT adjustments to the PD system. If it were curved or corded, it would be difficult to thread the bars through in the French system. Its weight will limit its use to short spans. A system that may compete with the PD would be a tub or short span box section which has deeper more effective shapes. The advantage of the PD system is its short profile which is advantageous for bridges with vertical clearance limits. Voids may help lighten the system. Marsh recommended keeping the top of the bottom flange of the PD roughened or it may curl or pull away from the CIP. Oliva thought that the PD system looked good and would be something that would be considered in WI. A void should be considered to reduce the shipping weight. Regarding a smooth surface at the top of the bottom flange, FHWA research has indicated that a smooth surface can help to avoid concrete fatigue issues. Because a crack will pre-exist or will develop at joints, designer should purposely design the system so that crack is likely to occur at location with higher reinforcement content. Prussack felt the PD was a robust system and had no doubt it would perform well. He thought the protruding rebar made things complicated in that forming would not be so easy, but he did not think it would be very expensive to build. He suggested avoiding voids. He thinks they are a pain to build. They need to be restrained to stay in place. Economically he said it would be cheaper to add a few more strands to carry the larger self weight than to use voids. Problems go up exponentially with size of voids. He stated every precaster has problems with them and often find that they move when casting. Rabbat commented that no matter how much CIP is required for the PD system, placing and finishing equipment would still be needed on site. A modification Rabbat suggested for the PD system would be to have one protruding hook a few inches above the other so the longitudinal bar could be slipped and

C-14 dropped into the joint which would make it easier for the longitudinal bars to be tied in place. [Although, note that currently, the longitudinal bars are dropped in place in a prefabricated cage.] Roberts-Wollman thought the PD system had promise for short span bridges. The thick CIP can help control reflective cracking. Potential issues of concern include differential shrinkage. Load induced cracking should not be a problem, but she expected that the system would crack. Ronald (HDR Engineered Structures) did not care for the PD system because of the field labor involved and the need to cast CIP in addition to providing precast. He did not feel that there was enough time savings with this system. He suggested considering PT with a threaded bar through the PD system. Russell liked the drop-in cage in the MnDOT implementation of the PD system. He commented on the closeness of the horizontal shear reinforcement to the surface (i.e., difficult to get concrete underneath). He also commented that most designers detail bars in the corner (i.e., extend S1609E closer to corner of concrete). He recommended concrete covers of 2 to 2½ in. on top and 1 in. on bottom from NCHRP study recommendations. He recommended going with LWAC if the CIP exceeds 8 in. in thickness. Seigurant thought that the PD systems may prove to be economical and assembled quickly as have the decked bulb-Ts. When given a choice between full depth deck systems and CIP, contractors often choose CIP as more economical. If speed of construction is an issue, the cost premium may be acceptable. Two comments on the system developed by MnDOT: 1) avoid stirrups protruding horizontally; enable vertical drop of the precast pieces into position; and 2) horizontal shear reinforcement should project out of the piece far enough to develop into the topping. Sennour felt the PD system looked promising, and would be a great solid span useful for county bridges. In response to those who think there is too much CIP in the PD system, he indicated that as long as CIP is being used, the thickness does not matter. He recommended a smooth top surface on the bottom flange as well as a flange that was as thin as possible. Smith did not care for the PD system, but thought it would be adequate for short spans. In terms of importance, he would place it at the low end of the spectrum. He thought it looked rigid and put a lot of dead load in the wrong places. Besides proposing voids, other suggestions Smith proposed from a fabricator’s perspective included using dowel bar stubs rather than having transverse bars protruding from the forms. The protruding bars can then be screwed into the couplers in the field. Tadros indicated the PD system had a limited place; he did not particularly care for it. He described the NE Inverted-T section which can span to 100 ft. He recommended avoiding situations where two torsionally stiff elements are connected. Anytime that is done, if they are not connected well enough, torsional cracks will develop (e.g., box beams). It is difficult to achieve longer spans with the PD system without a void (he recommended Chapter 8 of PCI Bridge Design Manual for guidance on design of transverse sections of adjacent boxes to make the deck width act as a plate). He recommended changing the top surfaces of the bottom flange to smooth. Tang really liked the PD and saw its potential on multispan and continuous bridges. He would like to see minimal CIP (just fill in voids between the webs and use precast top of web as riding surface). Also consider stronger joint connection details and LWAC. He was not sure if voids would work with the

C-15 system. He did not think that it mattered whether or not the top of the bottom flange was rough or smooth because any crack that would form should be arrested by the reinforcement. He recommended considering mechanical connectors as a means of connecting the reinforcement (e.g., screw-in or clip/swage together). Zendegui thought that rideability should be good with the PD system. He thought that the vertical bars in the PD system could result in many small cracks due to shrinkage resistance (may be better than one large crack). He also recommended avoiding the use of the bars protruding laterally, as they may cause forming problems (contractor refused to use them in their application because they did not want to damage the forms). C.4 Comments on Potential Means to Lighten the Precast Slab Span Superstructure Concepts A number of individuals surveyed indicated that means should be investigated to lighten the precast slab span system. Two options were mentioned: use of voids and lightweight aggregate concrete. This section summarizes these comments. Comments on use of voids Grafton (Pomeroy) indicated adding voids would be an option. In CA the ceiling is 48 kip on a flatbed truck (with up to 55 ft. length) without having to go to a more elaborate trucking unit. If the system is heavier than that, he suggests looking into the use of voids. If voids are required, look into using Styrofoam or cardboard. Corrugated metal can be used for voids up to 18-20 in. in diameter. If moisture gets into the voids, he suggests including a drain in the bottom and seals at the ends. If you can load the system on a truck without voids—avoid them. TX has good control over voids in forms; they sometimes pour in two sections [Hyzak]. WSDOT developed PD proposal with voids [Khalegi]. Culmo and Hawkins recommend looking into using voids if the system becomes deeper than 18 in. Issa built models with voids and tested them in the laboratory. He did not see a problem with constructing the PD with voids. Roberts-Wollman thought that voids could work if precasters can be convinced to use them. Tadros indicated that the PD system needed voids to span longer lengths. The NE Inverted-T section has voids made by topping the web with a timber form and casting the top flange. They have had no problem with floating voids with this construction concept and have experienced little leakage. The contractors also like it. Smith would recommend the use of voids to lighten the dead weight of the PD system. He is a fan of Tadros’ NE Inverted-T system; it is basic and easy to construct. Smith likes to patent products, and considered building Tadros’ system, but thought it would be difficult to patent. Ralls thought the PD system looked heavy without voids; however she indicated states like MD have changed from voided to solid systems due to voids popping up during casting. Dick opposed the use of voids as it increases complexity and adds new problems. MN and SD are opposed to voids because of previous durability issues experienced with voids.

C-16 Nickas did not recommend using voids because the forms float and end up costing more than a solid slab. Sennour also thought that voids could be problematic because of floating. If the system becomes very popular, he indicated that voids could be formed mechanically (as done in box beam systems). Russell thought that in general voids are fine; if they float then they were not tied down. Comments on use of LWAC Advantages that LWAC provides include cost savings on transportation and required crane needs. Grafton (Pomeroy) has used LWAC commercially but not for bridges. CA is interested in lightweight boxes for reduced superstructure mass and therefore reduced inertial forces created in response to seismic loading [Marx]. Although CalTrans is interested in LWAC, they have not yet used it that much (maybe less than 1 percent LWAC).WI is interested in lightweight. They have learned a lot over the last few years and think it looks promising [Hubbard]. Smith has had a lot of success with LWAC. Culmo associated with CT DOT and NH DOT is a big fan of LWAC. They used LWAC in SD in the ‘60’s and did not have problems with it, although cost may be an issue, Johnson thought it may be o.k. Prussack thought LWAC should be looked at as a tool, like magnesium phosphate grout. If it is possible to ship out more pieces with adjusted density of mix, he is for it. Back-calculate required density based on trucking limitations. Ralls indicated that individuals at TXDOT think that LWAC is the way to go for precast deck panels. She thinks that LWAC now has similar properties to NWC and good abrasion resistance. Breen at UT-Austin is doing some work in this area. Some DOTs have had disastrous results however. Lightweight concrete has performed poorly in AK and TX decks in past [Marx, Hyzak]. Marx indicated AK is hesitant because of poor experience in the past even though the problems were identified (i.e., reactive aggregate was detrimental to the mix). Material shortages have also been a problem in the past. TX would be interested in using it if research could provide success stories. IADOT has not had a reason to use it [McDonald]. Eriksson (USFS) had problems with LWAC in the past. Decks with estimated service lives of 20-25 years showed problems within 5 years. Castrodale, who is now with Stalite, indicated that LWAC is very viable. It has abrasion resistance that is suitable for a driving surface. It is also possible to get more uniform mix with LWAC compared with NWC. Many have had problems with LWAC in past, one of the lessons learned is the need to qualify the aggregate. Its reduced tensile capacity may lead to problems with the shear keys. Advantages include the ability to haul more pieces. It also can be used to span longer lengths. The aggregate/paste has better bond and with its lower stiffness is more forgiving than NWC and may lead to less cracking. Marsh suggested trying LWAC, but was concerned with aggregate segregation and lower strength of the material. He also noted that shear keys may not be as effective if made with LWAC. Russell and Sennour encouraged the use of lightweight materials in the project; Sennour uses it as often as possible. Abrasion tests indicate that LWAC is an acceptable material. The correct modulus needs to be used which affects the stiffness of the system including live load recovery and continuity. He does not think the freeze-thaw or chloride penetration issues are any more of a problem for LWAC than NSC.

C-17 C.5 Comments on Longitudinal and Transverse Connection Concepts between Precast Panels and Bulb-T Flanges Survey respondents provided some general comments on longitudinal and transverse joint connection concepts. These comments are summarized first, followed by comments specific to the five joint connection concepts identified in the survey: loop bar (U-bar) detail, straight bar detail with spiral to reduce lap length, headed bar detail, welded wire reinforcement (WWR) detail, and structural tube detail. These details are sometimes identified in the comments below as 5a-5e, respectively. The numbers correspond to the figure numbers in the handout distributed to the survey participants that described the connection concepts. C.5.1 General Comments on Longitudinal and Transverse Connection Concepts Culmo indicated that cross slope is always an issue in bridges that must be considered in developing connection concepts. CA is interested in the design of longitudinal and transverse jointed systems, along with the types of grouting available. Eriksson (USFS) thought all of the concepts looked more difficult than the weld tie system. Prussack is involved with NCHRP 12-69 with bulb-T flange connectors. Prussack and Rabbat would be interested in knowing how weld plate system would test out. Seigurant thinks that their welded connections with grouted shear keys work well. The weld plates also help correct for differential camber. With the connections shown, it is not clear how the leveling forces will be held until the grout/concrete achieves some level of strength, which is possible but will delay the introduction of traffic on the bridge. He thought that connections other than the loop bar detail are possible, but involve tight tolerances and difficult forming schemes. Grafton (Pomeroy) commented that most DOTs are used to CIP on prestressed I-girders. These jointed systems will be a new way of thinking for many DOTs. Ralls indicated the simpler the better regarding connection details. Issa thought that all of the proposed details would work. Zeldenrust thought some of the connections would require thicker decks (e.g., 10-11 in.); he would prefer 7½ in. decks. With higher dead loads, more girders may be needed. Kaufman (Morse Brothers Inc., OH) thought any reinforcement coming out of form would add cost. Straight bars sticking out of form (5b-e) create a safety hazard. Other than the looped bar detail and possibly the spiral detail, Oliva thought that the other connections would be difficult to form in the plant with the elements sticking out. Kowalsky thought the benefit of the spiral (5b), headed bar (5c) and steel tube (5e) was the rapidity of construction. For the transverse joint, he thought that the loop bar detail looked best followed by the spiral because of the capability to transfer moment. Marsh indicated that the joints may not provide the out-of-plane (vertical) shear capacity necessary to make the precast units share load, which needs further consideration before testing. Hawkins stated that it is desirable to have the reinforcement as close as possible to the surface for crack control and moment transfer, however joints may be lost quickly by corrosion attack. Koch (NC) indicated that the bulb-T system is a nice option to have, however, they would not use it on everyday bridges because they are expensive and less durable. It introduces joints that they have been

C-18 trying to eliminate for years. The place for these systems is with rapid construction. Ronald indicated that where there is differential camber, a haunch should be used on the girder. The flange-to-flange connection has too many drawbacks including the potential camber differential and large longitudinal joints. Henderson and Sletten have not used details like those proposed. They indicated that the connections are similar to closure pour details when the deck is cast on girders during phased construction. They have closure pours as narrow as 2ft.-6in. and recommend using mechanical connectors to develop the reinforcement. They suggested two websites for ideas: www.exodermic.com (contains details as to how their firm, LJB Engineers & Architects, have approached transverse and longitudinal joint details) and www.barsplice.com (for ideas of mechanical connectors). Perry thought that any joint detail would work with Ductal and not much space would be required in the joints because of the high bond strength and fluidity of the Ductal. Connections must be able to transfer moment if diaphragms are not used; Chandra recommended considering connection details used in cable-stayed bridges. Ericson with Splice Sleeve North America has experience with emulative details. He served as the chair of ACI 550 Precast Concrete Structures which produced report ACI 550.1R-01 on emulative detailing. Emulative detailing involves connecting adjacent precast segments bar-by-bar in concrete or grout, thus simulating CIP. Ericson discussed two systems that could be used to connect jointed precast elements: (1) NMB Splice-Sleeve® which is an oversized ductile cast iron tube which allows the rebar to fit inside even with minor misalignment; and (2) “Kwik Joint” developed by Al Yee. The bars are placed at middepth of the panel and dropped into couplers (open at the top) in adjacent panels. Tadros indicated that CIP should be avoided, because it slows down the construction. He recommended bolting things together to create a dry mechanical connection or a connection that is grouted later. Remove CIP from the critical path. If an overlay is included in the detailing, the CIP aspect is still included. The challenge is to execute the systems without CIP. I-beams with precast panels can be fit up with a built-in haunch and the top surface of the section could be ground if necessary. Potential barriers to the use of these systems where guidance is needed includes the 24 in. spacing of studs connecting panels to the girders [Volgyi]. It is difficult to judge the constructability of the proposed systems looking at the survey, would need to see field installation. Volgyi recommended getting contractor input. Badie recommended developing new connection concepts that do not protrude from the forms. C.5.2 Specific Comments on Five Originally Proposed Longitudinal and Transverse Connection Concepts The following summaries present comments from the respondents specific to the five connection concepts distributed to the survey respondents for their input. The general format of the summaries is to list the positives of the details, followed by the concerns.

C-19 Looped bar (U-bar) detail (5a) Bhide commented that the loop bar detail system has been successful in Japan [although in correspondence with some Japanese colleague there was some concern expressed in using these systems in high seismic regions]. Tang indicated that the loop bar detail was specifically developed for seismic regions. Oliva mentioned a similar loop system developed in Yugoslavia for seismic design in ‘80’s for wall system and wall/floor joint system. With this system you can force the crack through the overlapped reinforcement. A negative is that it requires a depth of deck which can be a problem for bulb-Ts. It may be possible to look into varying the deck thickness, but this may cause problems with differences in camber and sweep. Eriksson thought that this detail should be investigated further, and indicated that it would be the most seismically capable. Issa noted that the Koreans are also using a similar system, and he thought it was a good option. Chandra, Johnson, and Ma believed that this detail will work. Scott (NH) indicated that the looped bar detail was the one that interested him; Volgyi thought the detail looked “bullet proof.” Marx (AK) and Prussack felt this detail was the most promising, but Marx thought it may require a thicker deck which would be a negative in terms of added weight in transporting members, particularly bulb-Ts. Prussack noted that a similar detail was used in shear wall panels to link walls; he also saw some advantages to a thicker deck because it is hard to fit the loop bar detail into a bulb-T flange. Prussack did not think forming should be an issue, once the fabricator understands the detail, they can gear up. Prussack did not prefer use of stainless steel, he preferred a thicker deck (e.g., 8 in. think) or smaller bar (e.g., No. 4) to accomplish the detail. Tadros indicated the key to this detail was to get a waiver on the bend radius. He also mentioned wire can have a tighter bend radius because of the lower carbon content. Zendegui though that the loop bar detail looked promising for the use in the top of girders and for panel-to-panel connections. Forms can be built around the protruding loops. He thought potential problems would be with leveling and getting the shear connection. In addition, he thought there was too much steel inside the loops. Falhalla mentioned that they have had a similar detail in NE which has been in service for three years and has performed perfectly. He emphasized the concrete connection must be either nonshrink or expansive to ensure cracking does not occur and to avoid allowing moisture into the joint. In their project they required a Type-K cement, he recommends specifying material in a more generic way (e.g., expansive cement). Perry thought that the loop bar detail would work fine using Ductal. Nickas thought the looped bar detail was pretty standard. He thought that the loops did not need to be side by side, that they required staggering. Constructability of the system is fairly routine, and the top and bottom longitudinal bars are not included. FL contractors have lots of experience with this detail and recommend it. Ralls thought the detail looked attractive and indicated TXDOT is using something similar without the lip on the shear key. Seigurant mentioned that they have used a similar detail to join wall panels in precast floating structures and it has worked well. He thought that durability of the system can be improved if the joint is enlarged (as shown in Fig. 5a) and filled with concrete instead of grout. Sennour (CEG) indicated that the loop bar detail is used often in vertical joints of wall panels as a seismic detail He had a complaint with the rendering in terms of the number and size of bars. It looks very labor intensive. With a 6 in. bend radius, should not exceed No. 4 or No. 5 bar. In addition, he thought that 2 in. clear cover for precast should be plenty. Smith had experience with the loop bar detail, and thought

C-20 that it performed well. For the loop bar detail shown in the survey (Fig. 5a) he was more concerned with spalling chunks of concrete from the thin sections that butt; construction ease is not worth the risk. Culmo indicated that this connection would be the best detail for a moment connection in the longitudinal and transverse directions; however others questioned its capability. Rabbat questioned the ability of the looped bar detail to transfer moment. He thought it looked more like a hinge. He was concerned that there was very little straight bar to develop stress before the bend; however, he urged testing of this detail. Also Koch and Ronald were not certain about the transfer of moment without testing the detail. Miller thought this detail was simple, but when looking at positive moment development, must be able to mesh bars. Trying to position girders and get bars to mesh in, would be complicated. He has done work with bent bars and bent strands and believes that the bent prestressed strands worked best. The problem with the bent bars was that the bars could not be prebent before placing in the field. Several respondents were concerned with threading the longitudinal bars through and size of closure pour [Camp, Hyzak, Hubbard, Khalegi, Koch, Marsh]; other than constructability issue Culmo and McDonald (IADOT) indicated the detail looked promising. Culmo questioned whether longitudinal bars are needed in loops. R. Eriksson, on the other hand, indicated that must ensure enough longitudinal reinforcement within the loop to make the noncontact lap splice of the hoops effective. Tang thought the detail had potential constructability issues with the longitudinal bars. He was also concerned with the thin flange sections that abut and may have Q/C issues in handling/transport. In addition, radius of Japanese shear key poses constructability issues, he recommended straight edge with grooved interface. He preferred opening up the detail as was done in Mn/DOT implementation of PD. Kaufman was concerned with how to deal with skew with this system. In addition, he, Dick (PCI), Marsh saw a potential concern with breaking off flange tip (see Shape of Keyway section for proposed modification); Johnson (SD) did not have a concern with spalling off the flange tip. Ronald had experience with the loop bar system with mixed success. Unless it is constructed well, this joint will be the weak link; problems include loops not fitting right, interference, longitudinal bars not fitting through the loops. He also noticed the system, as proposed (5a), had potential for spalling off the thin flange under the joint. As noted above, Ronald also did not believe the detail would provide full moment transfer. Badie (NCHRP 12-65) is concerned with the tight bend radius of the system. To achieve tight bends, he believed the bars would have to be heated, which would increase the cost. Grafton (Pomeroy) indicated that the loop bar detail can be somewhat painful with the forms. It should be kept standard (e.g., at 6 or 8 in.) or as multiple (e.g., 4 in. spacing where unused holes could be taped). NYDOT used the loop bar detail but found it to be a fabrication nightmare in the precast plant [Flynn]. Part of the problem occurred because it was not in the contract to draw up details for the fabricator. There have been noticeable problems with the system in the field. Bonstedt thought that fabrication would be easy (sufficient development length, sturdy, forms easily on bottom), but constructability would be complex. Castrodale, Dorgan, Eriksson (USFS), and Zeldenrust indicated that it would be hard to fit the looped bar detail in the thickness of the deck; Castrodale and R. Eriksson indicated one way to fit the looped bar detail in the thickness of the deck would be to tilt the bars—but this would increase problems with fabrication including dropping the precast pieces in place. Eriksson

C-21 (USFS) also indicated the longitudinal bars would need to be flexible to fit into the system. Roberts- Wollman thought that the loop bar detail was a “pain” to use while trying to maintain the cover that DOTs require. She liked the idea of using stainless steel as an alternative but cautioned against the use of dissimilar metals. Straight bar with spiral reinforcement to reduce lap length detail (5b) Khalegi of WSDOT and Hawkins favored this detail of the jointed systems. Falhalla, Ma, Marsh, Prussack thought that this detail, as the loop bar detail (5a), looked like a good positive connection. Zendegui thought that the spiral detail, as well as the headed bar and WWR details, would work out fine for transverse connections. One concern is potential alignment problems. Issa favored this detail, but thought it would be more costly. Ronald also thought the detail looked expensive just to get a shorter lap splice; he also did not like the additional field labor required and potential alignment issues. Miller liked this detail, but thought it would require a large joint and would be expensive. Castrodale and Oliva thought the spiral detail may be o.k. Looks reasonable if spiral is small enough to compress easily for construction [Hyzak]. Stretching the spring would be labor intensive [Hubbard]. Johnson (SD), Koch (NC), Nickas (FL) thought that the system looked complex with the coil. Ericson indicated that this system was similar to Al Yee’s initial design to reduce splice lengths by 1/3 (predecessor to NMB splice sleeve). The system can be constructed by sliding the spirals to one side (need to be flexible). Mechanical connector is the next step beyond the spiral if it is required to reduce the width of the closure pour. McDonald (IADOT) indicated the connection looked reasonable but would have higher field labor costs than the other details. Badie thinks that this is a reasonable connection. He personally fabricated 3-4 panels with spirally confined joints at UNE and had no problems. Bonstedt thought that this system looked similar to one proposed by Tadros. His comments were that it would be difficult to construct, congested, and expensive (requires a lot of steel). Tadros indicated that this detail works very well and has been used previously in NE. He envisions the spirals would come to the jobsite precompressed and slid on the bars on one side of the precast. After the pieces are set, the spirals would be released and extended across the spliced reinforcement. Others with reservations about the splice detail included Hartmann (FHWA), Smith, and Sennour with comments related to the detail being congested and difficult to fabricate. Camp of NM and Dorgan of Mn/DOT thought that placement of the spiral would be difficult. Culmo thought this detail was complicated because it would have to be formed—he thinks PT could be done faster. Russell thought the detail was too labor intensive. Perry thought that the spiral detail would be redundant with Ductal, because the fiber in Ductal already provides confinement. Ralls indicated TX awarded a project in August ’04 with a similar detail. Roberts- Wollman indicated they attempted to use a similar detail on a precast panel—she suggested discussing it with contractors.

C-22 Headed bar detail (5c) Several respondents thought the headed bar detail had promise and preferred it over some of the other details. Bonstedt preferred this option over the spiral detail. Hubbard preferred it to the loop bar detail. Russell liked the headed bar and WWR details the best. Issa thought this system looked good, and this system like the WWR detail, would come to the field nearly completed. Sennour thought the headed bar detail was ideal (second only to the WWR if it could work), but would need testing. Marsh liked this detail better than the WWR detail; he thought that there would be more positive transfer for in-plane shear. Shear transfer, transverse to load is needed to hold the section together for flexure. For shear transfer in the vertical plane, the shape of the shear key is important. Dorgan thought this system might be the easiest to fabricate, but would have to be tested. As for the spiral and WWR details, Zendegui thought that the headed bar detail would work out fine for transverse connections. Falhalla has not had much experience with headed bars in NE, but thought that the detail looked fine. Ma (CA) and Nickas (FL) thought it could be used to reduce the development length, and consequently joint width. Others who thought the detail may have some advantages included Johnson (SD) and McDonald (IADOT). A few respondents indicated that his connection looked similar to one investigated in 12-69 [Hyzak, Prussack]. Culmo thought the detail would work but would be difficult to fabricate. Marx (AK) thought that both this and the WWR detail would be hard to line up. Camp, Khalegi and Koch thought it would be difficult to place the longitudinal steel under the headed bars; however other than that, Koch thought that this detail was the simplest. Castrodale thought it may be possible to tie the longitudinal bars to the underside of the headed bar on one panel and then position them in the field. Hartmann was concerned with the legs of the shear key potentially breaking off. Miller and Roberts-Wollman questioned whether this system could get the necessary development; Roberts-Wollman thought that otherwise it looked like a good detail “theoretically.” Miller thought the headed bar detail might be expensive. Welded wire reinforcement (WWR) detail (5d) Of the respondents who commented on this detail, Hubbard, Johnson, and Prussack indicated that it looked like it would be rapid to construct and doable. Bonstedt, Koch, and Nickas also thought that this connection would work. Issa also thought this system looked good, and this system like the headed bar detail, would come to the field nearly completed. Roberts-Wollman and Ronald thought the WWR detail had merit; Ronald thought it was more “fool proof” than the other details. Russell liked the WWR and headed bar details the best. Dorgan (Mn/DOT) thought the detail may be fine but emphasized that welding in the field should be avoided [note WWR would not require field welding]. As for the spiral and headed-bar details, Zendegui thought that the WWR detail would work out fine for transverse connections. Culmo thought it would work but would be difficult to fabricate. Sennour was uncertain about the WWR detail. It would need testing and he had a difficult time believing that the development is achieved and the section satisfies AASHTO requirements. If these are not factors, this would be his first choice because of the ease of constructability. Miller thought this detail may not be able to be fully developed. In addition to having trouble developing capacity; Camp, Hyzak and McDonald thought it would be difficult to fit two layers in the deck thickness. Marx (AK) thought

C-23 that both this and the headed bar detail would be hard to line up. Hartmann was concerned with the legs of the shear key potentially breaking off with this detail as well as in the cases of the other concepts. In addition, he had concerns with potential for fatigue, but was not certain whether that suspicion was justified. Ma indicated that WWR was not allowed in CA; he indicated that it may be a fatigue issue. He mentioned the work of Tadros on WWR, and although CA does not allow it now, manufacturers are pushing it. Marsh (Berger/ABAM) has never been enamored with WWR; it takes two cross wires to develop the tensile wire. Perry thought that it would be better to overlap alternating protruding bars and then drop in longitudinal bars rather than having the perpendicular bars coming from precast sections. Structural tube detail (5e) Koch (NC) thought contractors would like this detail the most of all. Marsh thought this detail looked “bomb proof,” provided there would be enough space; reflective cracking potential above the steel tube should be investigated. Dick (PCI) also thought the detail looked robust. Hubbard indicated that the detail looked unique and expensive [Hubbard]. Culmo thought the detail was worth consideration. Camp thought that this system required testing to ensure that the transfer of bar forces would be achieved. Perry and Ronald liked the tube detail; the heads on the bars should be large enough. Tang liked the tube detail and suggested an “L-shaped” slot to keep the bars in place, and noted that it would be difficult to place this detail in narrow joints. Sennour liked the tube detail; he thought this mechanical connection looked much better than trying to fit the bars through the loop detail; however tolerance may be an issue. He indicated that because DOTs typically have their own quality control personnel monitoring the fabrication of their products, it may be possible to achieve better tolerances. He indicated that one might expect that ½ to ¾ in. tolerances may be required, whereas for commercial use, attainable tolerances might be on the order of 2 in. A few respondents indicated this detail looked similar to one investigated in 12-65 [Hyzak, Ma]. Dorgan (Mn/DOT) and Bondstedt thought the system looked simple but may have alignment issues; Roberts-Wollman (VATech) and Hartmann (FHWA) were also concerned with alignment/constructability issues. Zendegui thought that this detail should have reinforcement running perpendicular to the tube and was concerned with alignment issues. Johnson liked this detail less than the others. Zeldenrust thought the tube detail would require tight control; it looks interesting, but difficult. McDonald (IADOT) and Nickas (FL) saw potential issues with sloppy field work and tube not filled with grout. Nickas also thought that the system had potential for lots of cracking before it engaged the bars. Russell thought that the detail would work, but if bars the bars were epoxy-coated, the tube would need to be coated. He was concerned with cracking with a whole row of these tubes within the joint. Badie indicated that the tube in Figure 5e looked too thick. Castrodale indicated that the detail may result in reflective cracking with the very stiff spot created by the tube. Bonstedt and Miller thought this system would be expensive and more difficult to place. Prussack did not care for this detail.

C-24 C.5.3 Comments on Common Issues of Concern with Connection Concepts Spalling of lip of joint (5a) Ronald, Smith, and Tang had concerns that the thin lip of concrete forming the bottom of the joint may spall; Smith stated that construction ease associated with butting the adjacent flanges is not worth the risk. Hartmann was also concerned that the concrete tips might crack off at the bottom of the shear keys in several of the other connection concepts, as well. Hubbard indicated that such spalling does not happen too much and WI would not have fear of it; their biggest concern would be rust prying off the piece. Johnson (SD) and Prussack also did not have a concern with spalling of the tip. Marsh indicated that keyways have performed well in the field. In cases where they have broken, he attributed it to large overloads. He noted that the joints of untopped panels leak occasionally, especially if the grout mix design, joint preparation, or grout placement are not done correctly. Form below joint in field A number of respondents indicated that requiring a form below joint in field should be avoided [Hubbard, WI, Prussack, Ronald]. R. Eriksson suggested that forming should be minimized. Ronald stated that anything that requires labor in the field adds cost and is problematic; all important quality control issues should be addressed in the fabricating plant. Ralls suggests minimizing closure joint width. However, rideability is a huge issue (especially for motorcyclists); if joint is 18 in. wide it may be possible to smooth out differences in elevation in adjacent pieces. Russell would like to keep the two interfaces as close together as possible. C.6 Comments on New Materials Many states are open to the use of new materials. Camp of NM is open to grouts with fibers, FRP and stainless steel bars. Prussack does not see much motivation for stainless and FRP because he has not seen problems with current systems and the new systems are costly. TXDOT is looking at HPC, corrosion inhibitors, SCC, FRP decks, LWAC [Ralls]. The following is a summary of the comments associated with a number of different proposed new materials. C.6.1 High Performance Steels (e.g., Stainless Steel, MMFX) Meinheit (WJE) recommended either epoxy-coated or stainless-steel clad rebar for use in CIP joints in precast decks to minimize or eliminate corrosion problems. He also suggested looking into ZBar stainless steel clad bars (www.gerdauameristeel.com). Volgyi indicated VADOT is leaning toward using stainless steel or MMFX. They have had lots of problems with epoxy-coated bar where the coating debonds. They initially used epoxy-coating on the top mat, but two years after implementation, they used it on both layers. NYDOT uses epoxy coated bar, but galvanized steel or stainless steel are options. Within the last year, NYDOT made change to require both top and bottom mat of steel to be epoxy-coated. Prior to that just the top layer was coated. NYDOT uses stainless steel where there is extra concern about durability

C-25 or if longer life is needed, or if where it would be costly to shut down the bridge. AK has used galvanized steel and had discussions regarding stainless clad steel. They are not opposed to trying new materials; the downside is the cost, Nickas concurred. WI has done two stainless clad projects but they needed to get material from England [Hubbard]. MMFX is more locally available. Stainless or MMFX may be more acceptable than FRP because they fit with current code requirements [Hubbard]. Castrodale thought both stainless steel and MMFX were reasonable options. Culmo, McDonald (IADOT), Koch (NC), and Johnson (SDDOT) were concerned with cost of stainless steel; McDonald and Tadros were also concerned with the lack of ductility of MMFX. Although Bonstedt indicated that he did not have much experience with stainless steel bars, he thought they would be harder to bend. He recommended investigating new materials to target 100 year design life. With current conventional materials (e.g., epoxy coated bars, HPC, SCC, etc.) may only be able to get 60-70 yr. life. Culmo had issues with the cost of these materials. Russell thought that stainless steel would be a good material to try. Zeldenrust (WSDOT) was not opposed to stainless steel or A706. Scott (NH) likes stainless steel, but has not seen recent cost estimates. Dick would consider stainless steel but indicated cracks would still be a problem in the long term for freeze-thaw issues. Ericson (Splice Sleeve North America) indicated that with stainless steel clad, roller is shallow and rounded which affects development in a deleterious way. Eriksson indicated that stainless steel has high elongations and suggested looking at Pfeiffer’s research from WJE. Hawkins indicated that it was a good idea to use new materials (e.g., stainless steel or MMFX). The FHWA IBRC program enables innovative things and one of the most common has involved use of stainless steel. Ronald did not think that stainless steel was necessary for this project. The introduction of stainless steel would mean the problem is unique to this specific precast system; as such it would not be a good job of emulation. Dissimilar metals In exploring the possibility of using stainless steel to achieve tighter bend radii or using the tube detail for the connection, several respondents raised concern that the use of dissimilar metals in the system could lead to the promotion of corrosion cells. The following is a brief summary of the comments. Meinheit (WJE) and Russell cautioned about the use of dissimilar metals as there would be danger of setting up a corrosion cell. Tadros stated dissimilar metals should be avoided. WI would have a concern with dissimilar metals [Hubbard]; they had a problem with single top layer of epoxy bars corroding. Ericson with Splice Sleeve North America would not recommend combining dissimilar metals because of the potential for a galvanic reaction. MacDonald indicated that when dissimilar metals are in contact it causes a problem; he indicated that if they were not in contact, he did not think it would be a problem. Schokker also commented on the potential to develop corrosion cells when using dissimilar materials in the joint (i.e., grout) with respect to the adjacent precast concrete panels; this is discussed further in Section C.6.3.

C-26 C.6.2 Fiber Reinforced Polymer (FRP) Reinforcement Bhide recommended looking at nonmetallic reinforcement in the systems. Russell thought that FRP was worth looking into but was unsure about the ability to bend it. WI is very interested in FRP. They have used grids and bars [Hubbard]. In NC there are two regions where corrosion is a concern, and in those regions, GFRP has given them good results. They have also used FRP wrap, but it is difficult for the inspectors to know what to look for in those cases. FRP is an option for SD, however they have not had much experience with it yet. Eriksson thinks FRPs are an option even though they are more expensive. Sometimes contractors just need to get used to new materials and with experience they may be strong proponents. Scott (NH) indicated that NH has built a bridge sing FRP grids made by Autocon in Ontario. Goodspeed oversaw the design, the deck is bare and there is no cracking; he indicated that “the bridge is beautiful.” He also noted that Goodspeed considered using epoxy in the joints on the project. TX has built bridges with FRP structural members and reinforcement; Hyzak expressed concerns with cost and flexibility. Culmo also had issues with the cost. Blaszak with TechFab is a structural engineer with many years of experience with FRP. He recommended considering FRP grid in place of WWR detail (5d). Advantages of the FRP include reduced cover and good crack control. He does not believe FRPs would be an option for the loop bar detail (5a). Dick would consider FRPs but as the case for stainless steel, he indicated cracks would still be a problem in the long term for freeze-thaw issues. Castrodale indicated that FRP may have stiffness issues. Oliva concurred. Upon cracking, FRP bars tend to be “softer” due to the reduced modulus. As a consequence there is likely to be more deterioration at the cracks with these bars [Oliva]. Flynn (NY) is interested in FRP, however the system cannot be designed in the same way. Because of the brittle nature of FRP, there is not a one-for-one swap with conventional rebar. Hartmann was concerned with out-of-plane distortion of FRP and fatigue. Marsh thought that the joints would not be wide enough to accommodate the development of FRP. Henderson and Sletten did not think that any of the connection details proposed looked like potential candidates for using FRP. They believe that the emphasis should be on grout materials. Sennour (CEG) strongly believed in exploring FRP for reinforcement—it is the way of the future. C.6.3 Grouts This section summarizes comments on materials to use within the connection regions. Most of the comments were on grouts, however, Badie and others also recommend investigating the use of ultra high performance concrete (UHPC) to reduce the development length of the reinforcement in the joints. Smith indicated that HPC was a good idea, but stated that it would not work for overnight construction. Smith also noted that it is not desirable to have a large percentage of deck area made with grout as it is difficult to get it done right and achieve good durability; therefore, minimize the use of grout. Ronald indicated that grout materials must be cost effective. The girder haunch material must be pumped in place and consequently needs a long pot life and must be flowable. New cementitious grouts (still very expensive) are being developed for PT that are just short of going to zero slump concrete.

C-27 Gulyas (former chair of ACI 223 Shrinkage-Compensating Concrete) has had a great deal of experience with grouts and epoxies including magnesium ammonium phosphate. Because something is called “nonshrink” does not mean it will not change dimension. Nonshrink grouts are evaluated in a sealed environment and “nonshrink” refers to the grout not undergoing settlement issues where water and vapor rise to the top and cause shortening. “Nonshrink” grouts are still susceptible to drying shrinkage. Cementitious grouts must be kept continuously wet to keep them from shrinking which is impossible. The use of less Portland cement and more aggregate is desirable, but it still does not stop shrinkage. Gulyas indicated one means of reducing the amount of shrinkage is to control the rate and amount of ettringite formed. Gulyas also mentioned interaction of reinforcement and grout related to restraint issues. Grout materials that do not shrink (as opposed to Portland cement grouts) include magnesium phosphate. It is a rapid set grout and can be used over a wide range of temperatures. If magnesium phosphate is used, surface preparation is very important (i.e., cannot have carbonation at interface). Some concern was expressed with the pot life of magnesium phosphate [Hyzak]. It may be sensitive to proportioning and mix temperature. Contractors may wish to work with grouts that are more forgiving (e.g., nonshrink cementitious grouts). Ericson with Splice Sleeve of North America indicated that if the grout layer is small (~½ in. to ¾ in.), sand would be used in the system and in the case of coupling columns, the grout layer is required to be ~2ksi stronger than the connecting elements. If the grout layer is thicker (e.g., 2 to 4 in.) pea gravel would be used in the mix. In Badie’s project, they are using Dayton Superior product (SS mortar) similar to that used in NMB Splice Sleeve. With this high performance grout they are able to reduce the anchorage lengths of their connection reinforcement. They estimate development lengths using ACI 318-05 equations assuming confined concrete compressive strengths. To avoid cracking in grout, MacDonald recommended maximizing the use of fly ash or slag. With the use of these materials, he indicated that cracks that would occur due to shrinkage could be plugged or would self-heal. Grout specifications should include a limitation on shrinkage (0.025 percent grout) or w/c max (0.220-0.225). He also indicated that there are a number of effective, but expensive, shrinkage reducing admixtures that reduce the surface tension of the water and thus reduce shrinkage. MacDonald proposed another option of filling the joints with an expansive material like Type A cement and add fiber to provide restraint. There is a 40 year track record of its use. Organic VMAs stabilize bleeding on a microscopic scale and have good results with reducing shrinkage. Tests to consider include ASTM C157 for shrinkage or ASTM C845 for expansive materials. Also look at permeability of the materials without fibers. Consider also freeze-thaw testing (although, if air entrained, it should not be a problem). Other materials to consider include self consolidating concrete (perform both U-box and L- box, filling and passing ability tests, respectively). Meinheit (WJE) developed specifications for the concrete mix used in the Wacker Driver reconstruction in Chicago. The role of WJE included designing the specifications for the concrete, conducting QC testing of the concrete during placement, and developing a method for fogging and blanketing after placement. The mix is a dense quaternary cementitious mix to avoid problems with freeze-thaw deterioration and alkali aggregate reactivity. He supports the use of shrinkage-compensated grouts (or grouts with controlled expansion). An alternative is to pre-place the aggregate and then pump the grout; in this case mechanical interlock between aggregate particles prevents global shrinkage. Fibers in the grout may

C-28 also be an option. Meinheit strongly cautions against using grouts with significantly different properties from the adjacent precast concrete (e.g., modulus of elasticity); however, he sees advantages to polymer modified grouts because they provide better bond to hardened concrete surfaces. Miller stated epoxy grout is not thermally compatible with the girder and will start peeling off the end of the girder. Another problem is a chemical solvent (e.g., Methyl ethyl ketone (MEK)) is needed to clean grout off surfaces, which may create personal and environmental safety concerns. Miller has used every type of non-shrink grout available, some bridges get cracking, some do not. His experience with nonshrink grout leads him to believe that non-shrink grout is rarely placed correctly. ODOT has had problems with placement of grout. Just because it comes in a bag does not mean it is easy to use. He suggests using a grout mix that can be delivered in a ready mix truck. Perry described the advantages of the Ductal system which is a high strength ultra high performance concrete. It might be used to reduce the joint widths. Ductal can also be dyed to any color to match precast systems. Perry indicated that is also good in seismic performance. It is strong and has shown good results in fatigue tests. Roberts-Wollman indicated that Ductal shrinks a lot; she indicated that using Ductal with FRP is a possibility. Regarding fibers in the grouts, Gulyas indicated that they tend to stick up and are not great for tires as concrete wears away. He suggested putting on metallic “dry shake” and troweling surface to bury the fibers. Marx (AK) indicated that fibers would not help resist a crack that would form at the interface between the grout and the precast. AK has used high performance prepacked nonshrink grouts including Set 45 [Marx]. They have also used epoxy-based grout supplied by ThermalChem. Flynn (NY) indicated that they use corrosion inhibitors (e.g., calcium nitrate in their grouts). They also seal the system with silane sealers. It is difficult to stop the closure pours from cracking, but they have had some success with a nonshrink grout with an elastomeric seal on top. Preventing moisture from getting into the grout has helped a lot. ThermalChem has experience in producing grouts (e.g., Product 3). By their very nature of being epoxy (i.e., nonhydrating material) they do not shrink. Product 3 also has a higher compressive and tensile strength than concrete. The biggest issue is the potential difference in coefficient of thermal expansion between the grout and adjacent material. This difference can cause the concrete to break around patches. They recommend mixing sand with the epoxy. A pourable system is a 4½ :1 ratio by weight (sand:epoxy). Temperatures can play havoc with the viscosity. Prussack indicated owner/agency handles grout specifications. He provides input along the lines of ASBI or provides specifications if asked (e.g., require high quality, prepackaged, non-shrink grout, with specified minimum strength of 4ksi and 5-6ksi as desirable range). They used to add aluminum powder (~2tsp) which provided non-shrink grout characteristics due to gaseous reaction of aluminum powder. Prussack indicates that non-shrink grouts today are vastly superior. Roberts-Wollman did research on grouts. They constructed 12x12x4 in. thick slabs with hole in middle filled with grout. The surface of the hole was moistened before filling with grout. The slabs were ponded regularly and after 6 mos., four of eight joints had cracks and started leaking. She did not necessarily think the problem was associated with shrinkage or cohesion. She suggested trying grout with pea

C-29 gravel instead of cement material. Adding fibers to the grout might help reduce cracking at places other than the interface. High flowability of the grout is sometimes important. Schokker stated that CIP must have similar properties to precast or a corrosion cell could be created. If the joint is more durable than precast and salts and water permeate the precast, major corrosion problems can occur. She stated that polymer or epoxy grouts should be avoided and suggested reviewing repair literature on polymer and epoxy grouts. Schokker recommended investigating different grout materials to show different consequences of not using durable materials. She also thought that freeze-thaw issues should probably be examined. High flowability grout may be needed. If corrosion/durability test procedures are proposed for use in the project, Schokker expressed willingness to review the proposed tests. Schokker indicated that expansive grouts should be investigated to counteract settlement, but indicated that if it is an additive, rather than bagged material, caution and high degree of quality control would be needed at site. Sennour believed in exploring high strength grouts; he thought it would be desirable to move the potential for cracking out of the joint region and into the concrete (consider two part sand to one part epoxy). Regular grout makes the joint the weak zone. Also consider shrinkage compensating or nonshrink grout that is Portland cement based. Shape of keyway A number of respondents commented on the importance of the shape of the keyway. Hyzak indicated the keyway shape shown in Figure C.6.1 would be effective in accommodating the transmission of compression struts across the joint. Marsh indicated that larger loads may require a sawtooth connection which could engage more struts. Tang was concerned with the shape of the keyway with the Japanese detail (5a). He suggested using a straight edge for the keyway with a grooved interface. Figure C.6.1: Effective shape of keyway Miller indicated that through FEM, he determined that full-depth shear keys are better than partial depth shear keys. For full-depth keys, the principle stress under load becomes longitudinal, while for a partial depth shear key, it becomes transverse. Ideally, Miller stated that a shear key would not be

C-30 poured all the way to the top, but would have a sealant on top; however, this then becomes a maintenance issue. Kaufman’s suggestion was to provide a ¾ in. chamfer, then a minimum 1 in. vertical, then start the shear key. Bhide indicated that the project should investigate whether there was a significant advantage to wider CIP joints as opposed to traditional female-female shear pocket grouted joints. Preparation of joint prior to grouting Nickas (FL) indicated that when replacing an old concrete deck, it is so dry, it drains the moisture from the new concrete. Consequently they require 4 hours of soaking before new deck is placed on top. On precast SIP, they require 12 hours of sprinkling with soaker hoses prior to placing the CIP. In TX, they usually pressure wash rather than sand blast the joints. They get the precast surface saturated surface dry (SSD) prior to grouting or use appropriate bonding agents [Hyzak]. Gulyas mentioned that hardened concrete should be presoaked or surface should be coated with epoxy (on dry concrete) so water cannot leave grout into the hardened material. If this procedure is not followed could experience curling. If epoxy is used, the epoxy should not be “too cured” prior to placement or could exacerbate formation of a crack. Good bond and well distributed reinforcement will distribute the stresses and control cracking. Prussack (Central Pre-Mix Prestress) emphasized the importance of surface preparation. Before a precast element leaves the plant, the outer surfaces (where form was attached) are sandblasted to remove surface paste and to create a texture for enhancement of bond. The precast is brought to SSD before grouting. The grout is then poured and cured with spray film or wet burlap. Prussack was not convinced that magnesium phosphate grouts work that much better than regular Portland cement based grout. The magnesium phosphate grout is more tolerant of cold weather and cures and sets faster—so it has an advantage in situations where bridge would have to otherwise be heated. MacDonald indicated that as long as the precast is clean, the CIP should bond with the precast. Sandblasting is good, but preferred close to time of placement of CIP. Wetting the surface without standing water is also effective. Culmo recommended sandblasting the keyways ahead of time and only using prepackaged grouts. Grafton (Pomeroy) indicated the shear key is sandblasted in the plant to take off form residue. Curing Keeping the time between casting and curing short to alleviate shrinkage was recommended. Hyzak also commented that raking the surface increases the surface area which causes even great moisture loss. Grinding In FL and TX, sections are made ½ in. thicker to accommodate grinding. Hubbard indicated that approximately ¼ in. should be ground from surface to accommodate roadways with faster speeds;

C-31 rideability would be problem without grinding. Koch (NC) prefers adding ½ in. extra cover to facilitate grinding, than adding asphalt. Grinding is not a problem, they typically grind and prior to putting on an overlay. Smith and Tadros did not have a problem with grinding. Tadros preferred it to CIP. Roberts- Wollman indicated that grinding and grooving are doable alternatives to overlays. WSDOT does not consider grinding an option because of reduction in potential cover. Dick suggested providing a surface that does not require grinding. Ralls also suggested avoiding grinding; it is done a lot in TX and is acceptable as long as extra cover is provided. C.7 Other Issues of Concern Related to Development of Connection Concepts This section summarizes additional comments that were provided related to the investigation of connection concepts. Some of the comments that were provided were general in nature. The following sections present the additional comments that are grouped into categories including differential camber, rideability, deck thickness, two layers of reinforcement, and skew. This is followed by a section of the more general comments. C.7.1 Differential Camber Bhide and Tang expressed concern with differential camber between pieces. Composite decks are used in PA to deal with any differential alignment issues; Bonstedt does not favor varying flange thicknesses to accommodate superelevations or connections at joints. R. Eriksson indicated that CIP overlays can cure a lot of problems such as differential cambers and cross slopes. Ronald indicated that in the case of differential camber, a haunch should be used on the girder. He thought the flange-to-flange system had too many drawbacks including the potential camber differential. Seigurant concurred that differential camber was the most concerning issue with respect to decked bulb-T construction. Koch (NC) did not think differential camber was much of a problem, but NC was wary about it. They installed inserts in the center of the girders at mid-span so that they could pull/pushdown the beams to match the panels. Smith in discussing tolerances indicated that joints ≥12 in. wide can also make up difference in camber, in addition grinding can be done if necessary. C.7.2 Superelevation Tang believes that proposed systems need to provide a means to accommodate crown and superelevation. Smith stated that for transverse panels, superelevation needs to be accommodated in a single plane. Thickening the deck is a “slippery slope.” It is not desirable for the bottom form to be on a different plane than the top plane. C.7.3 Rideability Rideability is an important issue for FL, NH, PA, and VA. Contractors are penalized in PA if the surface does not meet the International Roughness Index (IRI). Contractors sometimes need to grind even CIP

C-32 decks to meet contract obligations. Castrodale indicated that the ability to remove differential camber is important to rideability. Koch (NC) stated that PT and segmental bridges have lots of problems with rideability. Ronald indicated another drawback to the bulb-T, besides differential camber, is the complication associated with trying to achieve a smooth ride (e.g., profiling the surface of the bulb-T is heading in the wrong direction). Zendegui concurred that good rideability was difficult to achieve with deck bulb-Ts, but noted that the PD system would have good Rideability. NH prefers not to use bare decks [Scott]; however maintenance crews prefer them. C.7.4 Deck thickness The range in deck thicknesses commonly used in the different states varied. A few of the respondents provided information on the common deck thicknesses they use. CA uses 6.5 to 8 in. decks with a 1 to 3 in. asphalt overlay. FL uses a minimum of 8 ½ in. deck for CIP. MN uses 8 to 9 in. deck. PA always uses a composite deck, sometimes with SIP forms. The deck thickness is 6 to 8 in. depending on girder spacing. Zeldenrust (WSDOT) prefers a 7½ in. deck and expressed concern that some of the proposed connection concepts looked like they may require 10 to 11 in. thick decks. Eriksson (USFS) uses minimum deck thickness of 6 ½ in. including the ½ in. wearing course. For larger loads (i.e., U80) the deck is ½ in. thicker. A related issue is concrete cover (also see Section 3.6.3 for comments on grinding). For untopped panels, Marsh stated that it is desirable to have generous cover. Prussack indicated cover was traditionally 2 to 2 ½ in. on precast elements. C.7.5 Two layers of reinforcement in joint Regarding the ability to emulate CIP and to transfer moment effectively through the joint, several respondents commented on the issue of two layers of reinforcement in the joint. Johnson (SD) thought one would be fine, two would be better. Marsh (Berger/ABAM) thought that a single layer of steel for shear transfer is adequate, but he recommended two layers of steel for crack control and durability. In the case of a single layer of steel, if the joint is located over the web (e.g. precast deck panels with longitudinal joints), the single layer of steel should be skewed upward to provide negative moment capacity. Russell preferred two layers of reinforcement in the joint for bending resistance rather than a single layer near the middle. Prussack thought it may be difficult to get two layers of steel into many of the proposed details. C.7.6 Skew Some respondents indicated that connection concepts considered should be able to accommodate up to 30 degree skew; 90 percent of the bridges in FL have skews less than 30 degree. Hartmann concurred that skews up to 30 degree should be readily accommodated by the systems; he suggested that the panels be aligned normal to the beams and adjusted at the ends with special panels. Koch suggested investigating skews up to 45 degree. Tang recommended considering accommodating skews up to 15 to

C-33 20 degree; he indicated consideration should be given regarding how to place/connect panels (i.e., parallel to skew or use trapezoidal sections at ends). Curved bridges should be avoided. C.7.7 General Comments This section contains a number of general comments which did not fit into a specific category. They are organized here in alphabetical order of the respondents surveyed. Bhide thought that companion monolithic tests should be considered to validate the performance of the emulative systems. Marsh believed that it is more important to look in detail at new emulative systems rather than testing companion CIP systems. Need to define loads to be transferred, look at concentrated wheel effects. Also need tests for shear strength transfer in plane. Bhide also mentioned that issues related to design/detailing/construction of nonsymmetrical end pieces with integral crash barriers should be considered. Bondstedt indicated that one of the biggest problems with these precast systems is the crash barrier. There was a fatality in PA when a parapet broke off the bridge. If proposed systems do not corrode, can handle skew and superelevations, they would be sellable to PA DOT. Bonstedt also commented that there are regional issues to consider. In NE labor is expensive relative to materials, whereas in SW situation may be the opposite. For NE, innovations should be made on the precast elements to save labor in the field (and plant if possible). Chandra recommended the use of STM modeling to develop the details. He also recommended studying additional connections including foundation or pile connections to precast slab and column to cap connections. The investigation should also include a number of different ways of avoiding reflective cracking; ways of improving durability (e.g., SCC); different types of reinforcement; admixtures for precast and CIP closure joint to help with durability. He also indicated that in the past he had a problem with a microsilica overlay because the material had a higher modulus than the deck concrete; this was a problem in a long span structure where compression was introduced into the deck. Chandra also indicated the effect of traffic vibrations on green concrete may be of concern. Henderson and Sletten emphasized the importance of keeping water from collecting which would cause premature deterioration. Issa indicated that contractors do not like composite systems because they cannot make as much money when part of the cost of the project goes to the precaster, they prefer CIP. Khalegi of WSDOT suggests using anchor bars to fabricate system without protruding bars. MacDonald recommended considering performing a test of the bond between the concrete and the grout. Marsh recommended STM method for evaluating force transfer though the connection (see his comments on individual survey contained in Appendix B). He suggested that it may be possible to use a shear friction connection to transfer the shear in the connections; consequently, just need something to hold the joint together (i.e., compression from the diaphragm action or tension steel across the joint).

C-34 Ralls indicated UTDOT will use a new system as long as the cost does not exceed 30 percent more than the conventional method. IBRC/IBRD funds are a way to get innovative projects built. TX is interested in developing a system that can be constructed quickly with tolerance for camber mismatch. Their main concern is with durability. Russell suggested casting some of the specimens outside to investigate temperature effects on the joints. Schokker recommended investigating durability as it is a key issue. She indicated that good standard durability tests do not exist, but she recommended ponding saltwater on key connection details. Other issues that should be considered include those that drive the prices of precast include lifting weights and volume of use. McDonald, Tadros and several others strongly advocated for saving time in the field and doing most of work in precast plant. C.8 Issues to be Considered in Development of Guidelines This section summarizes some of the comments related to design issues and the development of guidelines. It includes general comments for consideration, as well as specific comments on issues such as restraint moments and bursting reinforcement requirements. C.8.1 General Comments for Consideration Eriksson recommends the use of detailed FEM models and STM calibrated with measurements to determine the capacity of the details. Simple rules for design (e.g., formulas, table) should be developed. Geometry will not vary much, so STM can be used for a parametric study to develop simple guidelines for practical design. Grafton (Pomeroy) suggested developing a catalog of standard details with provided design methodology. Hyzak suggested that an empirical method be provided for design. Once settled on a detail, it is convenient to apply it from project to project. Prescriptive specifications are desirable until achieve a comfort level with new systems [Hubbard]. Hartmann suggested relooking at bursting steel requirements for appropriateness. Also look at developing crack width limitations. Koch (NC) gave an example of useful specifications for full-depth precast deck slabs PT together. FHWA website provides prescriptive guidance on grouting and PT. Koch found this information helpful. Meinheit recommended that the selection and design of precast should strive for crack free superstructures for which the most effective technique is PT. As an alternative, cracks can be maintained tightly closed if sufficient reinforcing steel crosses potential crack surfaces. He recommends steel reinforcement twice the ACI (318-08) recommended minimums (i.e., at least twice 0.0018*Ag). This amount keeps cracks closed tight enough to minimize moisture penetration. Any cracks that form could be “self-repaired” by formation of calcium carbonate within the crack.

C-35 Ronald identified road blocks in the code to getting precast deck systems to be economical. These include AASHTO recommendations for cohesion and shear friction mechanisms for load transfer across the panel (e.g., shear studs placed at less than 2 ft. on center). In a bulb-T system with a 4 to 5 ft. wide flange, shear studs are not needed except at the ends where the rate of change of moment is high (i.e., maybe last 10 ft. of span) where shear is high. Other road blocks can be avoided by applying provisions of the code to a system that works. Areas where the code needs clarification includes required PT force between panels (e.g., 250 psi) to keep the joints from separating. An analysis should be done to determine if this amount of PT is necessary. In the case of joints that span transversely, the 250 psi may not be appropriate particularly where tension occurs. A 10 psi tension may not detrimentally affect the monolithic nature of the joint. Russell did not think that any specification covered the proposed systems adequately. The material obtained from this study would be useful in any format including AASHTO bridge specification, guide specification or guideline/drawing. He also recommended the development of a catalog of details as a starting point with suggestions for innovative opportunities. For initial installations, guidance on acceptable span lengths, specific requirements, are needed. Sennour stated that the main problem with bridge deck systems is that DOTs have gone to a “one design fits all” system (i.e., there is no variation in deck associated with different girder spacings). This limits the allowable girder spacing that can be used; girder spacing, deck thickness and strand number could be optimized to save money. Smith stated that guidelines on tolerances need to be developed. In addition, the design and bending capacity of decks connected with details such as the loop bar detail need to be determined. Tang suggested developing a catalog of 2-3 acceptable, standardized details. Provide a couple of options to deal with each different design situation. Try to eliminate interior joints on the deck, consider systems that are integral or semi-integral. Tadros indicated that there are specifications available for similar systems; however, they need improvement. Depending on the specification used (e.g., AASHTO Standard/LRFD, Ontario) different amounts of longitudinal steel are needed in the deck. Determine whether or not reinforcement needs to be distributed all over, is 250 psi minimum PT required, does shrinkage and temperature steel need to be distributed, is it additive to other reinforcement, etc. C.8.2 Restraint moments Restraint moments are an issue where CIP is cast onto a precast section. The moments are generated due to the differential movements (i.e., shrinkage and creep) between the composite materials. Hawkins said this issue was regional (e.g., Western Washington this is not a problem because of the moderate moist environment. Ma (CA) and Tang thought that restraint moments would not be a problem for shorter spans (i.e., situations where the PD system would be used). Oliva and Tang suggested specifying age of precast to avoid problems with restraint moments; Tang suggested 30 days as an example. Tadros also thought a waiting period would be ideal; otherwise he suggested that empirical details be developed rather than requiring calculations to deal with restrain

C-36 moment issues. Castrodale mentioned Miller’s NCHRP project (5-19) which included investigation of restraint moments. Miller’s work indicates that if a member is aged to 90 days, there is not an issue with restraint moments; whereas if the member is younger than 90 days, the designer needs to calculate restraint moments (see NCHRP 12-53). In AK sections sit for at least 90 days or more (sometimes 9 mos.) so restraint moments have not been a problem. AK is a zero tension state. If continuity for live load does not occur, they allow 3 to 6√f’c tension. Grafton thought that restraint moment issues could be avoided with letting the product age (7 days is too young, 90 days is excessive and would hurt the industry). Prussack concurred. He indicated most jobs are only 90 days from design to finish. Need a more forgiving option than waiting 90 days. NYDOT requires precast beams to be in storage for a minimum of 60 days prior to placement. They also use HPC for its qualities of lower shrinkage and creep. IADOT mainly uses prestressed beams with CIP pier diaphragms and restraint moments have not been an issue. Khalegi of WSDOT suggests extending strands into diaphragms to accommodate restraint moments (also helps for moment reversals due to seismic loading). WI has not considered restraint moments nor have they been concerned with restraint moment issues until 3-4 years ago with likelihood of change in specifications resulting from U of Cincinnati [Hubbard]. Systems (other than steel bridges) designed in TX as simply-supported, therefore they are not concerned with restraint moments [Hyzak]. R. Eriksson recommended addressing restraint moment issues and their transmission along the longitudinal joints. Consider using strands to establish resistance. Nickas prefers a simple solution (e.g., putting in reinforcement for worst case). C.8.3 Bursting Hawkins indicated that shear stresses being developed at end of beam need to be considered with respect to the development of bursting stresses. Sennour and Tadros saw no problem with bursting steel in deck panels. Culmo indicated that bursting is an area that could be addressed in AASHTO. Current requirements do not make sense for box beams or PD system, Oliva and Roberts-Wollman concurred. AK follows AASHTO Section 1301 which requires 4 percent reinforcement and 20 ksi. STM is used to model anchor regions. Tang stated that current specifications require too much reinforcement. This should be studied and reduced for appropriate situations. IA DOT uses debonded strands at ends of inverted T’s WI has not had problems [Hubbard]. WSDOT submitted a change to AASHTO relating to bursting steel requirements for a slab on a U-shaped member. C.8.4 Color between Joint and Precast Some of the respondents indicated that joints between the precast elements might create a driving hazard because drivers may tend to follow joint lines rather than striping. This section summarizes some of the comments on this issue. AK has not had a problem with joint lines. Koch, Smith, and Tang also did not see any drivers following joint lines as a major issue. Prussack indicated that he had never considered joints as confusing to drivers on well-traveled roads especially because of presence of fog lines and reflectors on busy roads. It does not seem that this has been an issue in Japan. Also deck surface may discolor from normal oils, etc.

C-37 In TX, they carefully tried to address color difference between grout and precast in Waco bridge. They were not successful. The color mismatch is noticeable and adversely affects safety—it may be necessary to use an asphalt overlay [Hyzak]. Falhalla concurred that it is important to require a uniform color so that drivers do not get confused. NB has required overlays so the appearance would be uniform, but would prefer being able to achieve a uniform appearance without an overlay. Falhalla thought it may be reasonable to strategically place the joints. Gulyas stated that grout will not match color of concrete but brooming in same direction as precast may help to hide the difference. In WS, drivers following grout joints rather than roadway striping is a concern. Falhalla of NB indicated that it is especially a problem in longer bridges. Hartmann concurred that he has heard of problems with drivers following joints, but he believes that it should be dealt with by others (e.g., DOTs). McDonald (IADOT) thought issue of joint lines confusing drivers was a minor issue. Bhide indicated that matching the joint color with the precast so the systems do not appear jointed should be added as a criterion (see above). Chandra also preferred matching the colors of the CIP to the precast. Mn/DOT has used overlays so this has not been an issue, Dorgan indicated that getting the precast system and grout to match would be ideal. Ma discourages the promotion of untopped diaphragms. He believes that it makes it difficult to distinguish lane lines. Sometimes decks are ground, but Ma prefers topping.