IRC:58-2011- A discussion

Although the IRC:58–2011 design method for concrete pavements was a major step

forward it still has many compelling issues at stake in the design aspects mainly due to so

much simplifications and assumptions in the application of versatile Mechanistic-Empirical

pavement design (MEPD) approach which otherwise has huge potential to provide an

optimize designed pavement structure. Succeeding paragraphs presents a detailed discussion

on all these issues.

1. Distress Prediction and Failure Criterion:

1.1 In IRC:58–2011, the rigid pavements are designed to resist both bottom-up as well as

top-down fatigue cracking damages. It was assumed that bottom-up fatigue cracking

will occur due to moving traffic loads during mid-day hours i.e. between 10 AM to 4

PM while top-down fatigue cracking will occur due to traffic movement during night

hours between 0 AM to 6 AM. Accordingly, a transfer function was used to evaluate

the allowable number of traffic load repetitions for each type of cracking mode or

mechanism i.e. bottom-up fatigue cracking and top-down fatigue cracking, separately.

Such approach is highly questionable due to following:

“The concrete slab is assumed to start with zero initial pavement damage (or

distress) immediately after construction stage and the damages due to each

subsequent load application are added to the existing damage. The pavement damage

due to each load application is calculated as a function of the ratio of applied load

stress to flexural strength of slab concrete. Using Miner's cumulative damage concept

(1945), pavement damage against assumed distress (say, fatigue cracking) is

expressed as a damage ratio between the predicted and the allowable number of load

repetitions. Fatigue damage in the concrete slab is related to cracking developed in it

as a result of either cracking mechanism or modes namely bottom-up fatigue cracking

or top-down fatigue cracking, depending upon the magnitude and sign (tension versus

compression) of the slab stresses. Both mechanisms give rise either to transverse

cracks, longitudinal cracks or corner cracks at critical (longitudinal or transverse)

edge in the concrete slab and consequently, resulting in fatigue cracking damage in

the concrete slab. Therefore, a coupling between these two mechanisms of fatigue

cracking is usually considered to relate the measurable physical distress (fatigue

cracking damage) with pavement response by a suitable transfer function.”

As such, separate consideration of fatigue damage by both cracking mechanism lead

to overestimation of the fatigue life (i.e. an increase in the number of axle repetitions

to failure) and consequently, result in unsafe design of the pavement structure and the

pavement structure designed using such design procedure may exhibit premature

failure due to fatigue cracking. Therefore, instead of using equation 2 (or equations 7

and 8 given on page 20 of IRC:58-2011), cumulative fatigue damage (CFD) should

be determined as under:

, ,

, ,

i j k

i j k

nCFD

N ≤ 1.0 (4.)

where,

ni, j, k = applied numbers of axle load repetitions for ith

axle load level for each jth

axle

type and kth

temperature gradient during specified period.

Ni, j, k = allowable numbers of axle load repetitions for ith

axle load level for each jth

axle type and kth

temperature gradient during specified period.

j = 2 (single axle and tandem axle) for bottom-up cracking and 3 (single axle, tandem

axle and tridem axle) for top-down cracking.

k = 2 (six hour mid-day period from 10 AM to 4 PM and another six hour night

period from 0 AM to 6 AM).

The above suggested approach is on similar lines as used in MEPDG 2008. Relevant

excerpts from MEPDG, 2008 are reproduced below:

“For JPCP transverse cracking, both bottom-up and top-down modes of cracking are

considered. Under typical service conditions, the potential for either mode of cracking is

present in all slabs. Any given slab may crack either from bottom-up or top-down, but not

both. Therefore, the predicted bottom-up and top-down cracking are not particularly

meaningful by themselves, and combined cracking must be determined, excluding the

possibility of both modes of cracking occurring on the same slab.”

Accordingly, accumulated fatigue damage after considering all critical factors for

JPCP transverse cracking is computed by the following general expression:

, , , , ,

, , , , ,

i j k l m n

i j k l m n

nFD

N ≤ 1.0 (5.)

where,

FD = total fatigue damage (top-down or bottom-up).

ni,j,k, ... = applied number of load applications at condition i, j, k, l, m, n.

Ni,j,k, … = allowable number of load applications at condition i, j, k, l, m, n.

i = age (accounts for change in PCC modulus of rupture, layer bond

condition,deterioration of shoulder LTE).

j = month (accounts for change in base and effective dynamic modulus of

subgradereaction).

k = axle type (single, tandem, and tridem for bottom-up cracking; short, medium,

and long wheelbase for top-down cracking).

l = load level (incremental load for each axle type).

m = temperature difference.

n = traffic path.

The applied number of load applications (ni,j,k,l,m,n) is the actual number of axle type k

of load level l that passed through traffic path n under each condition (age, season, and

temperature difference). The allowable number of load applications is the number of load

cycles at which fatigue failure is expected (corresponding to 50 percent slab cracking) and is

a function of the applied stress and PCC strength. The allowable number of load applications

is determined using the following fatigue model:

1.22

, , , , ,

, , , , ,

log 2.0 (6.)i

i j k l m n

i j k l m n

MRN

where,

Ni,j,k, ... = allowable number of load applications at condition i, j, k, l, m, n

MRi = PCC modulus of rupture at age i, psi

σi,j,k, ... = applied stress at condition i, j, k, l, m, n

1.2 Calibration and Validation: The (critical) pavement responses were correlated with

pavement performance indicators in the form of pre-defined pavement distress modes

for a given design life by empirically derived equations known as distress models or

transfer functions derived from the performance (or distress) prediction models based

on past experiences, field observations and laboratory results that compute the number

of repetitive loading cycles to specified pavement failure. Transfer functions are the

weakest part of any Mechanistic-Empirical pavement design approach. Therefore,

before applying to them in any pavement designs practice, they need to be calibrated

and validated to suit with local conditions in order to ensure a satisfactory agreement

between predicted and actual distress at project site otherwise it may result under-

designed or overdesigned pavement structure. This is the reason why emphatic

attention were given to carry out and verify the local calibration and validation of the

transfer functions in order to reduce the gap between predicted pavement distress and

in-service pavement performance[8 and 9]

. To locally calibrate and validate the distress

prediction model developed under NCHRP 1-37A in 2004, more than 50 research

studies are either completed or in near completion stage in USA, so far. In IRC:58–

2011, the transfer function is taken from Portland Cement Association) 1984 method

without any calibration and validation process in India. There is no published study or

reports which tell us how this transfer function is sensitive to varied conditions of

traffic, climate, material quality, mix designs, pavement constructions, and

maintenance practices etc. as per actual field conditions found in India? Due to this

only, design procedure will always be questionable. It is also worthwhile to note that

the details furnished in Appendix-III presents a comparison with the fatigue equation

used in IRC:58-2011 and not a validation as called therein. In fact, the objective of

validation is to demonstrate that the calibrated model of any distress mode can

produce robust and accurate predictions of pavement distress for cases other than

those used for model calibration. Validation typically requires an additional and

independent set of in-service pavement performance data. Successful model

validation requires that the bias and precision statistics of the model when applied to

the validation data set are similar to those obtained from model calibration. The

purpose of validation is to determine whether the calibrated conceptual model is a

reasonable representation of the real-world system, and if the desired accuracy or

correspondence exists between the model and the real-world system.

1.3 AASHO road test was conducted on rigid

pavements consisting either PCC slab + clayey

subgrade or PCC slab + GSB + clayey subgrade.

These pavement sections exhibit failure due to

erosion/pumping on account of erodible

subbase/subgrades. However, the relevance of

these results as said in clause 6.4.1 is difficult to

appreciate for the rigid pavements used on

National Highways in India, as shown in figure

1[1 & 2]

and keeping in view of the fact that 150

mm thick DLC layer (with cement content >7%)

used in India is extremely resistive to erosion.

Further, use of tied concrete shoulder further

reduces the propensity to erosion damage.

Figure 1 Concrete pavement used

on National Highways in India

It should be clearly specified, accordingly, in the guideline. Of course for all other cases, in

where DLC layer was not provided, this issue may become a vital cause of pavement failure.

Compacted subgrade

Natural Subgrade

GSB – drainage+filter layer (≥ 300 mm)

Concrete slab

(250 – 350 mm)

DLC slab (base layer) (150 mm)

1.4 In IRC:58–2011, pavement failure condition/distress criteria is not defined. Only one

structural distress mode of pavement failure (transverse fatigue cracking) is

considered. Other distress modes for pavement failures such as faulting, smoothness

etc. should also be made part of the IRC:58 in future. It is however emphasized that

failure criteria for each distress modes should be defined properly and user defined

distress criterion should be included in the design approach so that a designer can

design the pavement as per economy, importance and need of the project, and desired

level of pavement performance and adopted maintenance strategies.

1.5 In IRC:58–2011, commercial vehicles with the spacing between the front (steer) axle

and the first rear axle less than the spacing of transverse joints only are considered to

compute the top-down fatigue cracking damage against the consideration of all

commercial vehicles in MEPDG 2008. This will happen only when a structural model

capable to handle multiple (3, 5, 6 or 9) slab system with load transfer ability across

both transverse and longitudinal joints will be used.

2. Critical pavement responses: Currently, the critical stresses are computed at only

one or two locations for critical axle load placement. This approach is not correct for

multi-axle loads as the critical location is a function of the magnitude and

configuration of axle load, axle load placement, magnitude and sign of curling

stresses, load transfer efficiencies of transverse and longitudinal joints, construction

practices, and the pavement structure. To evaluate the maximum principal (design)

stress under any axle load location plus environmental loading, pavement response

should be evaluated at multiple locations on both longitudinal and transverse edges

and corresponding pavement damages (distresses) will be calculated for each location.

Location of maximum damage will be the critical location and should be made part of

design evaluation process.

3. Design of bonded JPCP: Design of a bonded concrete pavement structure is based on

the assumption of equating the flexural rigidity of concrete slab alone with that of

composite equivalent section of bonded slab and base layer for a fixed neutral axis

(NA) which ceases on considering the coupling between traffic load applications and

curling stresses considerations. For computation of pavement responses in case of JPCP

with or without doweled joints and with or without tied concrete shoulder finite element

method (FEM) are used while for bonded JPCP an approximate classical approach is

used. Such design and analysis approach is difficult to appreciate. In fact, finite element

method can be used for all JPCP cases having unbonded, bonded or partially bonded

interfaces between concrete slab and base layer through suitable structural modeling.

Further, interface bonding between PCC slab and base layer will break down after

certain period of service (5 years as per MEPDG, 2008) and therefore, assuming the

bonding state for entire design life may lead to an inadequate design.

4. Design of jointed plain concrete pavement with widened slab (by 0.50 m to 0.60

m) was assumed to be equivalent to that of jointed plain concrete pavement with tied

concrete shoulder. This assumption may be valid only when load transfer efficiency is

greater than 70% (against the provision of 50% for transverse joints and 40% for

longitudinal joints assumed in IRC:58-2011) otherwise widened slab condition always

results a saving of 2-5% in slab thickness besides superior performance in comparison

to any other alternative.

5. Characterization of Traffic: At present in India, there is no guideline to determine

equivalent number of single, tandem, and tridem axles for a commercial vehicle

classified in IRC:3-1983. It should be defined similar to FWHA. For example, for

class 9 vehicle, number of single axles is 1.13 and that of tandem axle is 1.93. For

vehicle class 10, number of single, tandem and tridem axle are 1.19, 1.09 and 0.89,

respectively. There is no meaning to assume it as mentioned in both examples of

IRC:58-2011. Further, guidelines for axle load distribution on Indian roads may also

be prepared. It is emphasized here that rigid pavement failure is more susceptible to

overloading generally observed in India (and which is not an issue of concern in

foreign studies) and therefore, determination of highest axle load’s share in total

volume is a vital issue for rigid pavements.

6. Structural modeling of the pavement structure: Choosing an appropriate structural

modeling of the rigid pavement structure is the first major step of Mechanistic-

Empirical pavement design approach to compute pavement responses in terms of

stresses, strains and deflections by finite element method. The modeling details,

analysis and design approach of the IITRIGID2 is not available. Therefore, it is not

known to the designer what will happen or in what way will he analysis and design

the pavement structure, if any of the input variables will vary? It is however appears

from the conspectus of the IRC 58 that single slab modeling with Winkler foundation

was used in IITRIGID2. Its comparison with other well established public software

such as EverFE, ISLAB2000, or KENSLAB etc. is required to verify its results and

validate IITRIGID2. Complete details of IITRIGID2 need to be spelt out elaborately

in an appendix for its further refinement. Design with IITRIGID2 needs to be

validated with MEPDG, 2008 software and RadiCAL, 2011. In any case, the

structural model should be capable of accurately predicting stress after duly

considering (a) the nonlinear temperature distribution in the concrete slab and

multiple wheel loads, (b) loss of support due to slab curling (separation of PCC slab

from foundation), (c) the effects of base course – the model must be able to consider

bonded, unbonded and partially bonded cases, and (d) slab-to-slab interaction in a

multiple (3, 5, 6 or 9) slab system and load transfer across both transverse and

longitudinal joints.

7. Temperature differential obtained in a CRRI study carried out in 1980s is still

considered in table 1 of IRC:58-2011 despite the large variation in climatic conditions

experienced now in the country and increase in the slab thickness used in the country

now-a-days. Further, it is much better to use a temperature gradient instead of fixed

temperature differential of -5 0C for accounting the in-built permanent curl in concrete

slab. Assumption of the fixed negative temperature differential is based probably on a

foreign study published in 1990 having different loading patterns, climate and

construction conditions which although can be used at the moment in the absence of

any Indian study but needs to be reviewed and investigated in view of local

conditions. Besides, other three components namely differential drying shrinkage,

moisture gradients, and creep may also be considered together with the in-built

permanent curl into an effective “built-in” curl.

8. Absence of Reliability concept: The design of rigid pavements is associated with

many factors that introduce a substantial measure of variability and uncertainties.

These factors include traffic prediction, material characterization and behavior

modeling, environmental conditions, construction quality, and maintenance practices

etc. Non-consideration of any load factor (as in 2002) in IRC:58-2011 further requires

that reliability concept is required to be introduced in the new IRC design practice as

a means of incorporating desired degree of certainty into the design process and to

ensure that the various design alternatives will result a pavement structure which

survive for the analysis period without reaching to unacceptable condition of

pavement performance.

9. The other glaring drawback of these design guidelines is that pavement design is still

applicable for a fixed set of conditions namely a standard rigid pavement structure,

slab dimension, vehicle characteristics, material properties of concrete, joint load

transfer efficiency at transverse joints and longitudinal edges, and design (failure)

criterion. Absence of any pavement design software is the biggest hurdle to perform

the pavement analysis and design with user-defined (or project-specific) input

variables and thus, to optimize the design and pavement structure. Sensitivity analysis

must invariably be included in the design guidelines.

10. Design of drainage layer: The net inflow (or quantity of water) removed by the

pavement drainage system must be determined from all sources that might contribute

to the possible saturation of the pavement section under consideration, including

groundwater, infiltration and melt water from thawing ice lenses (where frost action is

present) in the subgrade soil. In computing the net inflow, an allowance should be

made for any natural outflow that can take place by vertical seepage into the soil

beneath the pavement. Although crack infiltration rate for design purpose is

recommended as (2.4 ft3/day/ft) 0.223 m3/day/m, however, it can be suitably chosen

depending upon the frequency, intensity and duration of local precipitation. Assuming

the steady inflow uniformly distributed across the surface of the pavement section, the

thickness of drainage layer can be determined by applying the Darcy's Law (within its

application limits) as stipulated in IRC:58–2011 which result conservative design.

However, in case of non-steady flow condition (usually exists in the real pavement

structure), the thickness of drainage layer is determined by the following procedure

instead of using the Darcy's Law. Using the data of the example of drainage layer (pp

97-98) IRC:58–2011 as reproduced below:

Given

Pavement Infiltration rate (qi) = 0.120 m3/day/m

2 = 0.40 ft

3/day/ft

2

Resultant Length, LR = 16.24 m = 54.08 ft

Resultant Slope, SR = 0.039

Coefficient of Permeability, k = 333 m/day = 1109 ft/day

Determine required thickness of drainage layer (H).

Solution:

Figure 1 Chart for estimating maximum depth of low

(taken from ref. [11 & 12])

p = qi/k

= 0.40/1109 = 0.00036,

then from figure 1

LR/H = 185, hence,

H = 54.08/185 = 0.293 ft

= 87.7 mm.

or Alternatively, for given

thickness of drainage layer of

152 mm (0.50 ft), we have

LR/H = 54.08/0.50

= 108.16,

and SR = 0.039

p = qi/k = 0.00075,

hence,

k = 0.40/0.00075

= 533.3 ft/day = 160 m/day.

It is clearly evident that after assuming the non-steady flow condition, we can

economize the design of drainage layer. Further, instead of using the Crack Infiltration

approach, the Infiltration Ratio approach may also be included in the design process to

evaluate the amount of moisture infiltrating the pavement structure. Further, drainage

layer designed on the basis of depth-to-flow approach may also be checked for time-to-

drain method as proposed by Barber/Sawyer or Casagrande/Shannon. Software similar to

that of Drainage Requirements in Pavements (DRIP) Version 2.0 (2000) can also be

developed.

11. It is much better and practical to give the values of the coefficients of permeability for

all the coarse/open grades of GSB defined in MORTH specification or all the four

grades mentioned in table 8 of the IRC:15-2011 instead of quoting such values for the

gradation used in AASHTO 1993.

12. It is observed that the design thickness of concrete slab is almost insensitive to

variation in the effective modulus of reaction for foundation contrary to the fact that

slab thickness decreases by 2-4% on increasing the effective k-value for the

foundation (more stiff) from 50 MPa/m to 300 MPa/m. It needs to be clarified.

13. The thickness design of JPCP is observed to be insensitive with subgrade strength. It

is therefore suggested that the need to consider the reductions in subgrade strength

due to weak embankment beneath the subgrade will be considered only after the

confirmation by an Indian study.

14. Computation of the maximum bearing stress at the dowel-concrete interface is still

based on Friberg’s formula (1940) stipulated by equation (14), while assumption of

effective length equal to 1.0ℓ (instead of 1.80 ℓ as proposed by Friberg) over which

dowel bars are effective in load transfer across the joint was based on the results of

finite element modeling and analysis of dowel bars carried out by Tabatabaie et al.

(1979). Since then, a lot of research was take place for design of the dowel bars in

order to compute load transfer across joint, requirement and spacing of dowels, dowel

size to avoid the transverse joint faulting which leads to Mechanistic-empirical

approach. One such approach was used in MEPDG, 2008. Similar approach may also

be used for the design of tie bars which may also consider their looseness and

effectiveness in load transfer over the service life.

15. Thickness determination of either granular subbase or filter layer is very confusing.

IRC:58-2011 is silent on thickness design of GSB. The clause 6.3.3 of IRC:15-2011,

as reproduced below:

“….Grading types III and IV can be used at location experiencing heavy rainfall,

flooding etc. Cases where GSB is to be provided in two layers, it is recommended to

adopt grading III or IV for lower layer and grading I or grading II for upper layer.

Minimum compacted thickness of lower layer at locations where drainage

requirements are predominant shall not less than 300 mm. ….”

Now many questions are arises.

a. In which cases, GSB is provided in two layers? For such cases, what is the total

thickness of both layers of GSB, if lower layer is having a thickness of minimum

300 mm.

b. Whether such enormous thickness of GSB (>300 mm) is really required? If yes,

for what function?

c. It is however worthy to mention that the role of GSB in case of Indian rigid

pavement structure (where 150 mm DLC and doweled transverse joint are also

specified for any NH) is limited to act as drainage layer and of course, it should be

strong enough to sustain the construction traffic without any distortion/settlement.

d. For other cases, what is the design thickness of GSB?

Further, how a designer will determine the thickness of filter layer? IRC:SP:42 or

IRC:SP:50 gives the gradation requirement details of filter media around the edge drains.

Besides, Ministry’s specifications gives the filter design details provided under pitching in

case of slopes of guide bunds, training works and road embankments in which thickness of

filter media was specified as 200-300 mm. In my view, the thickness of filter layer may

specified in the range of 100-200 mm and that of drainage layer as 150-200 mm. Total

thickness of GSB (consisting both drainage layer + filter layer) should be less than 250-400

mm. Further, it should be specified that the filtration compatibility of the filter layer must be

evaluated with respect to both the subgrade and the drainage (subbase) layer to prevent

migration of the subgrade/filtration material into the subbase.

16. What is the worth of first three appendix of the IRC:58-2011 in spite of having too

long and heavy titles.

a. Except “a few” details of cement stabilized base in terms of strength requirement and

use of recycled material, no other details are given in Appendix-I. Are they suffice to

represent the title. Details given in IRC:74-1979, IRC:88-1984 and IRC:SP-89-2010

needs to be integrated and updated under one guideline which can be titled as

“Subbases and Bases used in Pavement Construction”. In this connection, one

wonderful document with title “Subgrades and Subbases for Concrete Pavements,

EB204P” published by American Concrete Pavement Association (ACPA) in 2007

can be referred.

b. Similar to Appendix “I’, a small paragraph containing incomplete details is suffice to

illustrate the said international practice on use of debonding layer over stabilized

subbase. As per my knowledge, two heavy spray applications of wax-based curing

compound are generally used for lean concrete subbases. Though there are no

common bond-breaker recommendations for stabilized subbase (either it is cement-

treated subbases or bitumen-treated subbases), various other alternatives of reducing

friction or bond between the concrete pavement and stabilized subbase exist,

including sand, bladed fines, asphalt emulsion, non-woven geotextiles, polyethylene

sheets, tar paper, and choker stone. Sometimes 25 mm thick dense graded bituminous

mix layer is also used. Application criteria, effectiveness of each alternative, ease of

availability and use, weather restrictions besides other parameters needs to be

included in this appendix to serve its intended purpose i.e. to improve the local

practices and make them as par with international practice.

c. Kindly refer to para 1.2 for comments on Appendix-III.

17. It may please be clarified what is the purpose of providing such a exhausting list of

references in IRC:58-2011 in contrast to earlier revisions if none of the reference

(except reference no. 6) is able to

a. demonstrate the advancement over the last decade as claimed in the introduction part

of IRC:58-2011,

b. illustrate or elaborate the design procedure adopted in IRC:58-2011, and

c. provide the reason(s) for the revision in IRC:58-2002.

Further, reference list includes only two studies from this vast country which have one

of the largest numbers of technocrats and researchers. It may please be clarified

whether both these studies are able to convey any meaningful result (it should be

remembered that one PhD study was completed in 1964 which have no relevance in

view of developments taken place over last 45 years and one M.Tech study has very

limited scope) utilized in the evolution of IRC:58-2011 and need to be included in a

National Specification.

18. It is also suggested that in a National Specification, write-up should always

introduced the subject, outlined the relevant assumptions for any design approach

along with application limitations, be specific, well referenced, and should never

concluding misleading facts – (last sentence of clause 3, clause 5.2, clause 5.6.2.1,

clause 6.4, clause 6.5.3, clause 6.7, clause 7.2.2, clause 7.2.5, table 5, and clause 7.2.7

needs to be reviewed). In any case, generalized and closed form approach should

always be avoided. A section briefing the limitation of earlier guidelines and reasons

for revisions must be included in the guideline. Examples furnished in the current

guideline may also illustrate this difference.