Guidelines for Dowel Alignment in Concrete Pavements (2009)

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.

38 The proposed guidelines are the recommendations of NCHRP Project 10-69 staff at the University of Minnesota. These guidelines have not been approved by NCHRP or any AASHTO committee or formally accepted for adoption by AASHTO. Introduction and Background Transverse joints are designed to allow slab movements due to shrinkage and thermal expansion/contraction while controlling the location and shape of slab cracks. Dowels are installed in these joints to improve load transfer capacity across the joints, thereby reducing slab deflections and stresses. Dowels must be properly sized and placed to carry applied loads and minimize longitudinal restraint (i.e., to allow joints to open and close, as needed) and they must be fabricated for durability (e.g., be resistant to corrosion or chemical attack). Dowels that are not located or oriented properly are called “misaligned” dowels. Misaligned dowels may not provide adequate load transfer capacity and/or may pre- vent the joint from opening and closing properly, resulting in premature pavement deterioration (e.g., joint faulting, spalling, etc.). Pavement dowels are generally installed using pre-fabricated baskets or cages (which are placed on grade before concrete placement) or by using a mechanical dowel bar inserter (mounted on the paving machine). Inspections of pavements in several states have shown that dowel misalignment generally occurs with both installation methods. These inspections have also shown that typical levels of misalignment do not always result in premature pavement distress. Requirements for dowel alignment were recently introduced based on limited in-service alignment and performance data (MTO, 2007; FHWA, 2007). The guidelines presented here are based on findings from field performance evaluation, laboratory testing, and analytical modeling and address the following topics: • Measuring dowel misalignment; • Quantifying the effects of misaligned dowels on pavement performance; • Determining critical levels of dowel misalignment that may result in lower levels of pavement performance; • Preventing dowel misalignment; and • Mitigating or remedying misaligned dowels in practice. Types and Definitions of Dowel Misalignment Dowel bars should be placed parallel to both the pavement surface and the longitudinal axis of the pavement in order to minimize longitudinal restraint of the transverse joints. Dowels are typically placed at mid-depth (to provide maximum shear load transfer capacity in the concrete slab) and the dowel bar should be centered longitudinally on the transverse joint. Misalignment is deviation in dowel placement from the pre- scribed position as a result of inaccurately placing the dowel, saw cutting in an incorrect position, dowel movement during the paving operation, or a combination of these factors. The five major categories of dowel misalignment, as illus- trated in Figure 1, are horizontal translation, longitudinal trans- lation, vertical translation, horizontal skew, and vertical tilt (Tayabji, 1986). Causes of Dowel Misalignment Common causes of dowel misalignment when using basket placement include: • Use of basket assemblies that are bent or are otherwise faulty due to inadequate rigidity (design), poor quality control during fabrication, or improper handling during transport and placement; • Failure to anchor the basket assembly to the grade prior to paving, thereby allowing the assembly to rotate, tip, or slide as the concrete is placed; A T T A C H M E N T A Recommended Guidelines for Dowel Alignment in Concrete Pavements

• Use of improperly sized basket assemblies (i.e., too tall or too short); • Mishandling dowels or baskets during concrete placement (e.g., workers stepping on dowel baskets); • Inappropriate basket or cage width or placement that inter- feres with slipform paver operation, resulting in rotation or sliding of the assembly; • Improper location of basket assembly; and • Improper location of sawed or formed joint. The most critical factor in using dowel basket assemblies is probably the number and type of pins used to secure the basket. When an insufficient number of pins (or inadequate pins) are used, the baskets may shift, rotate, or burst resulting in misalignment problems (ACPA, 2006). Common causes of misalignment when using dowel bar inserters (DBI) include: • Settlement of dowels in the concrete mass after insertion (due to mix fluidity, excessive vibration, etc.); • Movement of inserted dowels due to mishandling after placement; • Improper DBI operations; and • Improper location of sawed or formed joint. The most critical factor in maintaining dowel alignment when using DBI is probably the concrete mixture because it affects the ability of the DBI to accurately place dowels and control the dowel location and orientation in the plastic con- crete (ACPA, 2006). Detection/Measurement of Dowel Misalignment Many magnetometers, ground-penetrating radar units, and other devices can provide indications of dowel alignment with various degrees of accuracy. Because of the potential sensitivity of pavement performance to relatively small magnitudes of misalignment, measurement devices must be capable of pro- viding high precision. Among the devices used for measuring misalignment is the MIT Scan-2 (FHWA, 2007; FHWA, 2005; Yu and Khazanovich, 2005). Such measurement is performed by pulling the rail- mounted device along the joint while the device emits a weak, pulsating magnetic signal and detects the transient magnetic response signal. The included software which uses methods of tomography then determines the positions of the metal bars (ACPA, 2006). Available software can produce output in either numerical or graphical forms. A recent evaluation concluded that such a device measures dowel placement with an accuracy of ± 0.2 in. [5 mm] per 18 in. [457 mm] dowel length with 95 percent reliability on rotational alignment (FHWA, 2005). Typical Dowel Misalignments and Effects on Pavement Performance Dowel misalignment is expected to occur on every project. For example, variations in constructed slab thickness will result in variability in concrete cover over the dowels. Also, the accuracy of basket placement or insertion points during construction or joint sawing or forming operations will influ- ence embedment lengths. The following sections summa- rize the typical misalignment levels observed in the field in this study. Longitudinal Translation Field measurements indicated an average longitudinal dowel bar translation of 0.86 in. [22 mm], with project standard deviations ranging from 0.4 to 1.9 in. [9 to 49 mm], and 1.2 in. [30 mm] standard deviation for all individual dowels. The overall distribution of longitudinal translation measure- ments for individual dowels presented in Figure 2 shows that more than 91 percent of all bars measured were within 2 in. [51 mm] of being centered on the transverse joint and about 98 percent were within 3 in. [76 mm]. Vertical Translation Dowel bars are generally designed and assumed to be embed- ded at the mid-depth of the slab. Dowels that are closer to the pavement top surface are considered to have a negative vertical translation and those that are closer to the bottom surface are considered to have a positive vertical translation. The average absolute value of vertical translation for the individual dowels measured for a large number of dowels was 0.46 in. [12 mm] with a standard deviation of 0.6 in. [15 mm]. The distribution of vertical translations of individual dowel 39 Figure 1. Types of dowel misalignment (Tayabji, 1986). Horizontal translation Plan Horizontal skew Longitudinal translation Plan Plan Vertical translation Section Vertical tilt Section

bars shown in Figure 3 indicates that about 96 percent of all bars were within ±1.0 in. [±25 mm] of the mid-depth location; the remaining dowels were more than 1 in. [25 mm] closer to the top or bottom pavement surface. Dowel Rotation Dowel rotations about the horizontal and vertical axes (axial rotation is irrelevant for round dowels) are two forms of rotation that may significantly impact concrete pavement performance. Measurements on a large number of dowels indicated an average horizontal skew of 0.24 in. [6 mm] per 18 in. [457 mm] dowel with a standard deviation of 0.21 in. [5 mm]. Figure 4 presents the horizontal skew distribution for all dowels and shows that more than 60 percent of the bars were skewed by more than 1⁄4 in. [6 mm] per 18 in. [457 mm]. About 2 percent of the bars had horizontal skew values exceeding 0.75 in. [19 mm] and about 0.5 percent had horizontal skew values exceeding 1.0 in. [25 mm]. Vertical tilt averaged 0.23 in. [6 mm] per 18 in. [450 mm] dowel with a standard deviation of 0.21 in. [5.4 mm]. Figure 5 presents the vertical tilt distribution and shows a very similar distribution to that of the horizontal skew. About 9, 2, and 1% of dowel bars had vertical tilt more than 0.50 in. [12 mm], 0.75 in. [19 mm], and 1.0 in. [25 mm], respectively 40 0% 5% 10% 15% 20% 25% 0.0 to 0.25 in 0.25 to 0.5 in 0.5 to 0.75 in 0.75 to 1.0 in 1.0 to 1.25 in 1.25 to 1.5 in 1.5 to 1.75 in 1.75 to 2.0 in 2.0 to 2.25 in 2.25 to 2.5 in 2.5 to 2.75 in 2.75 to 3.0 in 3.0 in to 6.5 in Longitudinal Translation, in. Pe rc en t o f B ar s Figure 2. Distribution of longitudinal translation for field study measurements. 0% 5% 10% 15% 20% 25% 30% 35% < -1in. -1 to -0.5 in. -0.5 to 0.0 in. 0.0 to +0.5 in. +0.5 to 1.0 in. > 1.0 in. Pe rc en t o f S ec tio ns Vertical Depth Deviation, in. Figure 3. Distribution of vertical dowel bar translations. Figure 4. Distribution of the horizontal skew. 0% 10% 20% 30% 40% 50% 60% 70% < 0.25 in. 0.25 to 0.50 in. 0.50 to 0.75 in. 0.75 to 1.00 in. 1.00 to 1.25 in. 1.25 to 1.50 in. 1.50 to 3.50 in. Pe rc en t o f B ar s Horizontal Skew, in.

Summary of Misalignment Values An analysis of data from 60 project sites indicates that most joints had dowel misalignments within the following ranges: • Longitudinal translation: ±2 in. [±51 mm] over 18-in. [457-mm] dowels. • Vertical translation: ±0.5 in. [±13 mm] for pavements 12 in. [305 mm] or less in thickness. • Rotational components (horizontal skew and vertical tilt): each less than 0.5 in. [13 mm] over 18-in. [457-mm] dowels. Measures for Reducing Dowel Misalignment Several measures can be taken to reduce the potential for dowel misalignment, as discussed in the following sections. Design Issues • Dowel baskets should be designed to withstand the rigors of transport, handling, and placement. Baskets that are not sufficiently rigid may bend or allow the dowels to be moved during construction. • The use of narrower baskets, with the outside dowels located 9 to 12 in. [229 to 305 mm] from the pavement edge and longitudinal joint (instead of 6 in. [152 mm]) reduces the cost of the basket (one less dowel is used if spacing remains constant) and reduces the probability of the paver catching the dowel basket and shoving or twisting it during paving. Construction Issues • Proper care must be taken in the storing and handling of dowel baskets at the job site to prevent bending of the basket or misalignment of the dowels. All dowel baskets should be inspected prior to and during the paving process. Damaged baskets should be removed and replaced prior to placement of concrete. • The correct locations for dowels should be marked along both edges of the pavement for either basket or DBI place- ment methods. The marks must be placed accurately and must be easy for the saw crew to locate after the paver has passed (ACPA, 2005). • Dowel baskets must be accurately placed in proper alignment on the survey marks. A thorough inspection of basket and dowel alignment prior to paving is extremely important. • Dowel baskets must be firmly staked or anchored to ensure that they do not move or tip during paving. Low anchor points help to prevent shoving and sliding of the basket while high anchor points help to prevent tipping of the basket (ACPA, 2005). • The types of anchors used and the frequency of their use should be selected based on the type and thickness of base used. For example, a 6-in. [152-mm] pin may be used to firmly anchor the basket to an asphalt-treated base (ATB), but it may need to be installed on a skew if the layer thickness is 4 in. [102 mm]. • There is no consensus with regard to the treatment of dowel basket tie or spacer wires during construction. The FHWA recommends that these wires should be removed, citing concerns that failing to cut the wires may contribute to joint lockup and subsequent slab cracking and notes that prop- erly anchored baskets do not need these wires for stability (FHWA, 1990). The American Concrete Pavement Asso- ciation (ACPA) recommends that dowel basket tie wires should not be cut after basket placement and prior to paving because cutting the tie wires may destabilize the basket, allowing it to come apart during paving and result in mis- aligned dowels. ACPA also states that analyses show that concerns about the contribution of tie/spacer wires to joint lock-up and subsequent slab cracking are unfounded (ACPA, 2005). • Care must be taken during construction to avoid stepping on the dowel baskets and dowels, especially during paving. • When using a DBI, concrete mixtures should be selected to ensure stability of the dowel bars during placement and subsequent paving operations (i.e., vibration, screeding, etc.) • To eliminate possible confusions between tie bars that are placed between adjacent lanes and/or shoulders and dowel bars that are placed at transverse joints, tie bars should not be placed within 2 ft [0.6 m] of transverse joints. Misalignment Limitations While no clear relationship was found between moderate levels of dowel misalignment and pavement performance in terms of faulting, spalling or panel cracking, laboratory testing 41 Figure 5. Distribution of vertical tilt. 0% 10% 20% 30% 40% 50% 60% 70% < 0.25 in. 0.25 to 0.50 in. 0.50 to 0.75 in. 0.75 to 1.00 in. 1.00 to 1.25 in. 1.25 to 1.50 in. 1.50 to 4.00 in. Pe rc en t o f B ar s Vertical Tilt, in.

and analytical modeling determined that dowel misalignment could reduce dowel shear capacity and its ability to transfer a load. Therefore, limitations on dowel misalignment can be based upon their effects on load transfer effectiveness, and by the minimum acceptable concrete cover (with respect to either the top or bottom of the slab) or the depth of joint saw cuts. Dowel alignment guidelines and specifications should stip- ulate requirements that are achievable with good construction practices and have no significant adverse impact on pavement performance. However, specifications and guidelines should encourage placement that is as accurate as is reasonably pos- sible, and also recognize that certain levels of misalignment may not significantly affect pavement performance. The following approach will help establish dowel placement specifications: 1. Establish constructible acceptance criteria. Establishing a relatively tight (but constructible) placement tolerance will promote the placement of properly aligned dowel bars and eliminate the need for further evaluation or remedial actions. Examples of such tolerances may include the following: • Horizontal or vertical rotational alignment: 0.5 in. [13 mm] over 18.0 in. [457 mm]. • Vertical translation: + 0.5 in. [13 mm] for pavements 12 in. [305 mm] or less in thickness; + 1.0 in. [25 mm] for pavements greater than 12 in. [305 mm] in thickness. • Longitudinal translation: 2.1 in. [55 mm] over 18-in. [457 mm] dowels. • Horizontal translation: 1 in. [25 mm]. 2. Establish rejection criteria. Rejection criteria should be established on the basis of measured, predicted, or expected pavement performance or behavior. For example, remedial action may be required due to inadequate depth of place- ment (considering concrete cover requirements and saw cut depth), inability to achieve specified performance thresholds (e.g., predicted faulting or IRI), or obvious placement flaws (e.g., interference from misplaced tie bars). For example, a dowel, joint, or section may be rejected if any of the following conditions occur: • Concrete cover at any end of the dowel is 2 in. [51 mm] or less from the top surface. • Concrete cover from the dowel to the top surface is less than the sawcut depth. • Rotational misalignment is 3 in. [75 mm] or more per 18 in. [457 mm] dowel length. • Agency-specified performance prediction measures are not met. The following procedure may be used for analyzing the effects of dowel misalignment on the performance of a uniform pavement project: 1. Use dowel alignment measurements to calculate the equiv- alent dowel diameter in each joint of the pavement project, using the procedure described under “Joint Effectiveness Evaluation”. 2. Establish “uniform sections” approximately 500 ft [153 mm] in length for the purpose of analysis and evaluation such that all joints in the section have similar equivalent dowel diameters. A series of several joints in any 500-ft [153-mm] section with equivalent dowel diameters that are rela- tively uniform but substantially different from the rest of the section may be evaluated as a separate uniform section. 3. Compute the mean equivalent dowel diameter for each section. Two or more adjacent sections with no significant difference in equivalent dowel diameters can be combined into a single section for analysis purposes. 4. Perform MEPDG computations for each uniform section using the calculated mean equivalent dowel diameter for the section, and compare the performance and distress predictions for each section with the prescribed perfor- mance thresholds or the as-designed pavement performance prediction. A decision about the acceptance, rejection, correction factors, etc. can be made for each pavement section based on the results of the computations and stipulated threshold values. If correction measures (such as dowel retrofitting) are performed, the effective dowel diameters of the affected joints should be recalculated and the pavement performance pre- dictions for those joints reassessed. Corrective Measures The following corrective measures may be considered appropriate for dowels or doweled joints that fail to meet the acceptance criteria: • Dowel bar(s) with inadequate concrete cover or excessive rotational misalignment can be corrected by removing and replacing the misaligned bars (retrofitting). – Bar(s) in the wheel path that cannot be removed can be corrected by removing and replacing the entire joint using a doweled full-depth repair. – Bar(s) not in the wheel paths that cannot be removed can be corrected by cutting completely through. • Individual dowel bars with inadequate embedment or that are missing can be corrected by retrofitting additional dowels. • Average initial joint load transfer efficiency should not be less than 70 percent after corrective measures have been performed. 42

Joint Effectiveness Evaluation The equivalent dowel diameter concept assumes that a joint with misaligned dowels behaves as a joint with perfectly aligned dowels of a smaller effective diameter. The equivalent dowel diameter, deq, is defined by the following equation: where remb = adjustment factor for a reduction in embedment length; rcc = adjustment factor for a reduction in concrete cover; rvt = adjustment factor for vertical tilt (dowel rotation); rhs = adjustment factor for horizontal skew (dowel rota- tion); and d0 = nominal dowel diameter. The procedure for selecting appropriate adjustment factors for the different misalignment forms is described below. Embedment Effect (Longitudinal Translation) Figure 6 can be used to estimate the adjustment factor remb (or dowel diameter ratio), for dowel embedment length, Lemb. No reduction or adjustment is assumed for embedment of 6.9 in. [175 mm] or more, and embedment of 2 in. [51 mm] or less should be treated as undoweled (i.e., remb = 0). The relation- ship shown in Figure 6 is given by the following equation for embedment lengths between 2 and 6.9 in.: For example, the adjustment factor, remb, for an 18-in. [457-mm] dowel that is longitudinally translated by 3.9 in. [99 mm] (i.e., 5.1 in. [130 mm] of embedment), is 0.916. r L Lemb emb emb= − × ×0 010 0 167 0 3242. . . d r r r r deq emb cc vt hs= × × × × 0 Joint load transfer efficiency (LTE) is generally most affected by the dowel(s) closest to the applied load (generally in the wheel path), and the dowels in the wheel path affect LTE as much as the combined effects of all other dowels in the joint. The following procedure should be used if dowel embedment varies along the joint: 1. Compute the adjustment factor for each dowel in the joint. 2. Determine the mean adjustment factor for all of the dowels in the joint. 3. Determine the mean adjustment factor for the three dowels in the critical wheel path (for example, the right wheel path in the truck lane). 4. Use the average of the two values obtained in Steps 2 and 3 as the adjustment factor for the joint. Concrete Cover Effect (Vertical Translation) When the dowel is translated vertically, the amount of concrete cover is reduced either above or below the dowel, which reduces the shear load capacity of the concrete and of the dowel-concrete system. The reduced shear load capacity can be represented by a dowel diameter reduction factor for concrete cover, rcc, which is the ratio of the diameter of a dowel placed at mid-depth having the same shear capac- ity as the vertically translated dowel in question to that of the diameter of the misaligned dowel. The reduction in effective dowel diameter depends upon the amount of ver- tical translation, the typical amount of variation in vertical translation, and the assumed basic or reference amount of concrete cover. A reduction should be applied only when vertical translation exceeds normal variability and the result- ing concrete cover is lower than a specified reference level of concrete cover. The actual concrete cover (CC) for a given dowel can be calculated as follows: • CC = y – d0/2, where y is the measured depth of the center of the dowel and d0 is the nominal dowel diameter if the longitudinal axis of the dowel is above mid-depth of the pavement • CC = HPCC – y – d0/2, where HPCC is the slab thickness if the longitudinal axis of the dowel is below mid-depth of the pavement (i.e., the dowel is closer to the bottom of the slab). Alternatively, CC can be computed equivalently using the following single equation: The reference level of concrete cover, CCref, can be con- sidered as the amount of cover above which there is no CC H 2 d 2 H 2 yPCC 0 PCC= − − − 43 Figure 6. Adjustment factor for embedment length. 0.0 0.4 0.8 1.2 0 1 2 3 4 5 6 7 8 9 Co rre ct io n F ac to r (r e m b) Embedment Length, in.

appreciable increase in dowel-concrete shear capacity (or below which there is a decrease in dowel-concrete shear capacity). For any pavement thickness, the maximum possible concrete cover, CCmax, is developed when the dowel is located at the exact mid-depth of the slab: Because there is some expected variability in concrete cover due to variances in constructed slab thickness, finished base profile, etc., the maximum possible value of concrete cover should be reduced by this amount of variability to serve as a basic reference level of concrete cover for evaluating the effects of vertical dowel misalignment. Thus, one possible value of CCref is given as: where Δtyp = the typical variance in vertical dowel translation, which is estimated to be 0.5 in. (13 mm) for slab thickness of < 12 in. (305 mm) and 1.0 in. (25 mm) for slab thickness greater than 12 in (305 mm). Finite element analysis indicates that increasing concrete cover beyond approximately 3.5 times the diameter of the dowel bar does not improve the shear capacity of the dowel-concrete system. Therefore, the reference level of concrete cover can be defined as the lesser of the theoretical amount of concrete cover for a dowel at mid-depth of the slab less the typical vertical translation, or 3.5 d0. This is expressed as follows: If the actual CC is greater than or equal to the reference level of concrete cover (CCref), no reduction in dowel diameter should be considered (i.e., rcc = 1.0). If the actual CC is less than 2 in. [51 mm], then the adjustment factor should be assumed to be equal to 0 (i.e., rcc = 0). If the actual concrete cover is less than the reference value (but greater than 2 in. [51 mm]), a reduction in effective dowel diameter should be considered. Laboratory testing and finite element analyses of dowels embedded in 8-in. [203-mm] concrete slabs have provided the following relationship: rcc = − − ( ) + ( )+1 153 3 2503 153 32. .  CC CC CCref ref ( )⎡⎣ − ( ) ⎤⎦ 2 2503 9628 CC CC Min H 2 d 2 d for slab thic ref PCC 0= − −( )0 5 3 5 0. , . kness 12 in. 305 mm CC Min H 2 d 2ref PCC 0 ≤ [ ] = − − . 1. , .0 3 5 0d for slab thickness 12 in. 305 mm ( ) > [ ] . CC H 2 d 2 H dref PCC 0 typ PCC 0 typ= − − = −( ) −Δ Δ2 CC H 2 d 2 H dmax PCC 0 PCC 0= − = −( ) 2 For example, the adjustment factor due to vertical translation of a 1.5-in. [38-mm] dowel in a 11 in. [279-mm] thick concrete slab by 1.75 in. [44 mm] (i.e., concrete cover is decreased from a reference or “nominal” level of 4.25 in. [108 mm] to 3.0 in. [76 mm]) would be: The following procedure is recommended for determining the “average” effective diameter for dowels with variable concrete cover at a particular transverse joint: 1. Compute an adjustment factor for each dowel in the joint. 2. Determine the mean adjustment factor for all of the dowels in the joint. 3. Determine the mean adjustment factor for the three dowels in the critical wheel path (for example, the right wheel path in the truck lane). 4. Use the average of the two values obtained in Steps 2 and 3 as the adjustment factor for the examined joint. Rotational Effects (Vertical Tilt and Horizontal Skew) The effects of vertical tilt and horizontal skew have been observed to be similar in both laboratory tests and analytical modeling. Therefore, the two adjustment factors, rvt and rhs, are computed in the same manner but separately. Determining adjustment factors for vertical tilt and hori- zontal skew requires the measurement of vertical tilt and hori- zontal skew for each dowel in the joint. These data are used to compute the mean tilt or skew, standard deviation of the tilt or skew, and the maximum tilt or skew of the dowels in the crit- ical wheel path. This information can then be used to estimate the stiffness of the joint and the joint load transfer efficiency (LTE). LTE can be related to the average diameter of properly aligned dowels to compute rvt and rhs, as illustrated below. The following relationship between dowel tilt and non- dimesional joint stiffness was developed: where JStiff = nondimensional stiffness of a joint with rota- tionally misaligned dowels; JStiff0 = the predicted nondimensional stiffness of a joint with aligned dowels (see Table 1); MeanTilt = absolute value of the average tilt (or skew) of the dowels in the joint, in. per 18-in. dowel; JStiff JStiff MeanTilt StD= − × − ×0 0 20623 0 61796. . Tilt WPTilt− ×0 86862. rcc = − − + + − 1 153 3 4 25 2503 153 3 2502 2. . .  4.25 3.0 3 0 82  3.0 9628 = . 44

StDTilt = standard deviation of the tilt (or skew) of the dowels in the joint; and WPTilt = maximum absolute value of the tilt (or skew) of the dowels in the wheel path, in. per 18-in. dowel. Example: Assume that the average measured horizontal skew of 1.5-in. [38-mm] diameter dowels in a joint is 0.2 in. [5 mm], with a standard deviation of 0.633 in. [16.1 mm] and that the maximum horizontal skew of the wheel path dowel is 0.8 in. [20 mm]. The resulting nondimensional stiffness value would be 9.767. The nondimensional stiffness can be used to calculate LTE using the following relationship (Crovetti, 1994): Thus for a nondimensional stiffness value of 9.767, the LTE would be 85.2 percent. This value then can be used to deter- mine the adjustment factor due to rotational misalignment, rrot, as follows (0.98 in this example): Combined Effect The equivalent dowel diameter concept assumes that a joint with misaligned dowels behaves as a joint with perfectly aligned dowels of a smaller effective diameter, deq, as defined by the following equation: For the example illustrated above, the equivalent dowel diameter for the misaligned 1.5 in. dowels (assuming all of the misalignments occur concurrently and that there is no vertical tilt) would be: deq = × × × × =0 916 0 82 1 0 0 98 1 5 1 10. . . . . . in. d r r r r deq emb cc vt hs= × × × × 0 r d LTErot = ( )0 0103 0 0582 0 . exp . LTE JStiff = + ( )− 100 1 1 2 0 849 % . . Example for Calculating the Equivalent Dowel Diameter This example assumes an 11-in. [279-mm] thick pavement with joints containing 12 dowels with 18 in. [457 mm] length and 1.5 in. [38 mm] diameter, with the following conditions: 1. The saw cut was incorrectly made 4 in. [102 mm] away from the designed location, resulting in 4 in. [102 mm] of lon- gitudinal translation and 5 in. [137 mm] of embedment length for all 12 dowels. 2. The dowel basket was 0.75 in. [19 mm] taller than was required for the mid-depth dowel placement, resulting in 0.75 in. [19 mm] vertical translational displacement towards the pavement surface and reduced concrete cover for all 12 dowels in the joint from 4.75 to 4 in. [121 to 102 mm]. 3. The rotational misalignments (vertical tilt and horizontal skew) for all 12 dowels in the joint are given by Table 2. The dowels are numbered according to their distance from the truck lane shoulder (i.e., dowel Number 1 is the closest to the shoulder). The first three dowels are considered to be wheel path dowels. Calculation of Equivalent Dowel Diameter Embedment Length Adjustment Factor Because the embedment length is greater than 2 in. [51 mm] and less than 6.9 in. [175 mm], the adjustment factor due to the longitudinal translation and reduced embedment length, remb, is computed as: remb = − ( ) + ( )+ =0 010 5 0 167 5 0 324 0 9092. . . . r L Lemb emb emb= − + +0 01 0 167 0 3242. . .  45 Table 1. Nondimensional joint stiffness values (JStiff0) for aligned dowels of various dowel diameters. Dowel Diameter (in.) JStiff0 1 6.537 1.125 7.447 1.25 8.461 1.375 9.601 1.5 10.894 Table 2. Assumed dwell misalignments in the joint. Dowel Bar Number 1 2 3 4 5 6 7 8 9 10 11 12 -0.44 -0.50 -0.34 -0.80 -0.54 1.46 -0.54 0.46 -0.54 -0.54 -0.54 -0.54 -0.26 -0.32 -0.32 -0.38 -0.48 -0.27 -0.39 -0.33 -0.47 -0.43 -0.44 -0.42 Vertical tilt, in./18 in. Horiz. Skew, in./18 in.

Vertical Translation (Reduced Concrete Cover) Adjustment Factor The reference and actual concrete cover values (CCref and CC, respectively) are computed as: Using these values of and CC, the adjustment factor due to the loss in concrete cover, rcc, can be calculated as follows: Vertical Tilt Adjustment Factor For the vertical tilt measurements provided in Table 2, the following misalignment parameters are calculated: • Mean vertical tilt = 0.2 in. [5 mm]. • Standard deviation of vertical tilt = 0.633 in.[16 mm]. • Maximum absolute value of wheel path dowel vertical tilt = 0.5 in. [13 mm]. The nondimensional joint stiffness is calculated as follows: The LTE of the joint can be estimated as: The vertical tilt adjustment factor associated with this load transfer efficiency can then be estimated as: LTE = + ( ) =− 100 1 1 2 10 03 85 51 0 849 . . . % . LTE JStiff % . . ( ) = + ( )− 100 1 1 2 0 849 JStiff = − ×( )− ×(10 8942 0 20623 0 2 0 61796 0 633. . . . . ) − × =0 86862 0 5 10 03. . . JStiff JStiff MeanTilt StD= − × − ×0 0 20623 0 61796. . Tilt WPTilt− ×0 86862. rcc = − − ( ) + ( )+ ( )1 153 3 4 25 2503 4 25 153 3 42 2. . . .  ⎡⎣ − ( ) ⎤⎦ =2503 4 9628 0 968 . CC H d H 2 y in. PCC PCC= − − − = − − − = 0 2 11 2 1 5 2 11 2 4 75 4 . . .00 in. CCref = − −( )the smaller of H d and 3.5PCC 2 2 0 50 . d or 3.5 in. or 0 11 2 1 5 2 0 5 1 5 4 25 5 = − −( ) = . . . .  . .25 4 25in. in.= Horizontal Skew Adjustment Factor For the horizontal skew measurements provided in Table 2 the following misalignment parameters are calculated: • Mean horizontal skew = 0.38 in. [10 mm]. • Standard deviation of horizontal skew = 0.073 in. [2 mm]. • Maximum absolute value of wheel path dowel horizontal skew = 0.32 in. [8 mm]. The nondimensional joint stiffness can be calculated as: The LTE of the joint can be estimated as: The horizontal skew adjustment factor associated with this load transfer efficiency can then be estimated as: Because the maximum allowable adjustment factor cannot exceed 1.0, an adjustment factor of 1.0 will be assumed. Computation of Overall Effective Dowel Diameter The equivalent or effective dowel diameter is the original dowel diameter (d0) multiplied by the adjustment factors for concrete cover, embedment length, vertical tilt, and hori- zontal skew: rhs = ×( ) =0 0103 1 5 0 0582 85 98 1 02 . . exp . . . r d LTEhs = ( )0 0103 0 0582 0 . exp . LTE = + ( ) =− 100 1 1 2 10 49 85 98 0 849 . . . % . LTE JStiff % . . ( ) = + ( )− 100 1 1 2 0 849 JStiff = − ×( )− ×10 8942 0 20623 0 38 0 61796 0 073. . . . .( ) − × =0 86862 0 32 10 49. . . JStiff JStiff MeanTilt StD= − × − ×0 0 20623 0 61796. . Tilt WPTilt− ×0 86862. rvt = ×( ) =0 0103 1 5 0 0582 85 51 0 995 . . exp . . . r d LTEvt = ( )0 0103 0 0582 0 . exp . 46

Therefore, to account for the effects of the misalignment in this example, the pavement should be treated as if it had a dowel diameter of 1.31 in. [33 mm] (and not 1.5 in. [38 mm] diameter). Assessment of a Pavement Section Problem Statement The following example illustrates the calculation of the effect of dowel misalignment on the performance of a 540-ft. [165-m] pavement section with an 11 in. [279 mm] thickness. The pavement section has 30 joints, each of which contains d r r r r deq emb cc vt hs= × × × × = × × ×0 0 909 0 968 0 996 1. . . × = 1 5 1 31 . . in. 12 dowels with 18 in. [457 mm] length and 1.5 in. [38 mm] diameter. The pavement was designed with the following per- formance criteria after 20 years at 90 percent reliability: • Transverse cracking not to exceed 12% of cracked slabs. • Mean joint faulting not to exceed 0.12 in. [3 mm]. • IRI not to exceed 160 in./mile [2.5 m/km]. The equivalent dowel diameters were calculated for the dowel alignments of each joint; results are shown in Table 3. Because the pavement section is less than 1000 ft [305 m], the mean equivalent dowel diameter is computed for the entire pavement section resulting in 1.41 in. [36 mm]. This equivalent dowel diameter was then used in an MEPDG simulation to predict faulting and IRI for the project. Figures 7 and 8 present the predicted faulting and IRI, respectively, for the as-designed 47 Joint # Equivalent Dowel Diameter (in.) Joint # Equivalent Dowel Diameter (in.) Joint # Equivalent Dowel Diameter (in.) 1 1.31 11 1.5 21 1.21 2 1.5 12 1.22 22 1.5 3 1.41 13 1.5 23 1.5 4 1.14 14 1.5 24 1.27 5 1.5 15 1.49 25 1.5 6 1.1 16 1.5 26 1.5 7 1.5 17 1.5 27 1.5 8 1.5 18 1.23 28 1.5 9 1.5 19 1.05 29 1.37 10 1.5 20 1.5 30 1.5 Table 3. Equivalent dowel diameter for each joint in the pavement section. 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0 2 4 6 8 10 12 14 16 18 20 22 Fa u lti ng , in . Pavement age, years Performance Threshold 1.5 in. 1.41 in. Figure 7. Predicted faulting for the as-designed pavement project.

pavement (dowel diameter of 1.50 in. [38 mm]) and for a similar pavement with 1.41 in. [36 mm] dowels. The predicted faulting and IRI of the project after consid- ering the dowel misalignment effects are within the specified acceptance thresholds. However, analysis of the MEPDG run output files (not presented here) showed that because of dowel misalignment, the reliability of faulting not exceeding the performance threshold was reduced from 96.7 to 91.9%, and the IRI reliability was reduced from 92.5 to 91.0%. Concluding Remarks The guidelines provide a simple methodology to account for the effects of dowel misalignment in estimating pavement performance. The methodology uses an equivalent diameter concept in which dowel diameter is reduced to account for the effects of misalignment. The dowel diameter reduction factor depends on the type and level of misalignment. Equa- tions for determining the reduction factor for different types of misalignments were developed based on the results of field, laboratory, and finite element analysis. Pavement performance can then be estimated using pavement analysis predictions (e.g., the MEPDG) for the reduced dowel diameter. Attachment A References ACPA (2005). The Relationship Between Sawed Joints and Dowel Bars. Concrete Pavement Progress, Vol. 41, No. 3. American Concrete Pavement Association. Skokie, IL. March 30, 2005. ACPA (2006). Evaluating and Optimizing Dowel Bar Alignment. SR999P. American Concrete Pavement Association, Skokie, IL. ARA (2005). Dowel Bar Alignments of Typical In-Service Pavements. Publication SN2894. Portland Cement Association, Skokie, IL AASHTO (2008). Mechanistic-Empirical Pavement Design Guide: A Manual of Practice. Interim Edition. American Association of State Highway and Transportation Officials, Washington, DC. Burnham, T. (1999) A Field Study of PCC Joint Misalignment near Fergus Falls, Minnesota, Report No. MN/RC-1999-29, Final Report. Maplewood, MN: Minnesota DOT. Crovetti, J.A. (1994). Evaluation of Jointed Concrete Pavement Systems Incorporating Open-Graded Permeable Bases. Ph.D. dissertation, University of Illinois at Urbana-Champaign, Urbana, IL. FHWA (1990). Concrete Pavement Joints. Technical Advisory T 5040.30. Federal Highway Administration, Washington, DC. November 30, 1990. FHWA (2005). Use of Magnetic Tomography to Evaluate Dowel Bar Placement. TechBrief. Federal Highway Administration, Washing- ton, DC. FHWA (2007). Best Practices for Dowel Placement Tolerances. Tech- Brief. Federal Highway Administration, Washington, DC. Fowler, G., and W. Gulden (1983). Investigation of Location of Dowel Bars Placed by Mechanical Implantation, Georgia Department of Transportation. Report No. FHWA/RD-82/153. Federal Highway Administration, Washington, DC. Khazanovich, L., and A. Gotlif (2002). Evaluation of Joint and Crack Load Transfer. Final Report, FHWA-RD-02-088, Federal Highway Administration, Washington, DC. MTO (2007). Ontario Provincial Standards for Roads and Public Works. Sections 350.04 through 350.08. Ontario Ministry of Transportation, Toronto, ONT, Canada. Tayabji, S.D. (1986). Dowel Placement Tolerances for Concrete Pave- ments. In Transportation Research Record 1062, National Research Council, Washington, DC, pp. 47–54. Yu, H.T. (2005). Dowel Bar Alignments of Typical In-Service Pavements. R&D Serial No. 2894. Portland Cement Association, Skokie, IL. Yu, H. T., and L. Khazanovich (2005). Use of Magnetic Tomography Tech- nology to Evaluate Dowel Placement. Report No. FHWA-IF-06-006. Final Report. Federal Highway Administration, Washington, DC. 48 80 120 160 200 240 0 2 4 6 8 10 12 14 16 18 20 22 IR I, in . / m ile Pavement age, years Performance Threshold 1.41 in. 1.5 in. Figure 8. Predicted IRI for the as-designed pavement project.