Chapter 18 — Pavement Subgrade Evaluation

Pavement subgrade evaluation for railroad facilities encompasses two distinct but related infrastructure domains: (1) paved surfaces within railroad right-of-way — intermodal container yards, locomotive servicing facilities, maintenance-of-way (MOW) vehicle access roads, classification yard roadways, and fueling pads; and (2) the granular track substructure system — ballast, subballast, and subgrade — which functions as an engineered pavement system for track loading. Both domains require rigorous characterization of subgrade support capacity, compressibility, moisture sensitivity, and frost behavior.

This chapter provides methods for characterizing subgrade strength (CBR, M_r, k modulus), designing flexible and rigid pavement structural numbers using AASHTO 1993 methodology, evaluating frost penetration and heave potential for CPKC and CN northern operations, and developing subgrade improvement recommendations for intermodal yard heavy container crane trackage. All equations are presented in English units. Railroad-client-specific pavement standards are integrated throughout.

Railroad pavement subgrade evaluation differs from highway pavement design in several important respects:

Loading regime: Intermodal yards subject paved areas to heavy rubber-tired container handlers (reach stackers up to 180,000 lb gross weight; loaded highway tractors at 80,000 lb GVW), overhead gantry crane rail loads (250–500 kip axle loads on crane rail foundations), and stacked container loads (4-high stacking of 45,000-lb gross containers = 180,000 lb per stack pad). These loads are far more severe than typical highway loadings and require higher subgrade strength criteria.

Crane rail foundations: Container cranes at intermodal yards ride on embedded crane rails set in reinforced concrete beams spanning pile caps. The geotechnical input — differential settlement between crane rail beams — governs structural crane serviceability more than pavement structural number alone.

Frost sensitivity: CPKC and CN operate intermodal terminals and MOW facilities in areas subject to frost depths of 4–8 ft (Prairie provinces, Manitoba, Ontario, Quebec, British Columbia interior). Subgrade frost heave of even 1–2 in. creates unacceptable differential movement in crane rails and reach stacker lanes.

Heavy industrial use: Unlike highway pavements rated by equivalent single axle loads (ESALs) over a 20-year design life, railroad intermodal pavement is often designed for 15-million-ESAL equivalents to reflect 24/7 terminal operations with heavy handling equipment.

Client-specific standards table:

JPCP = Jointed Plain Concrete Pavement; HMA = Hot-Mix Asphalt; ESAL = Equivalent Single Axle Load.

Reference graphics

18.2.1 California Bearing Ratio (CBR)

The CBR test (ASTM D1883 — laboratory; ASTM D4429 — field) measures resistance of a compacted soil to penetration by a standard cylindrical piston at a rate of 0.05 in./min. The CBR value is expressed as a percentage of the standard force required to penetrate a standard aggregate:

CBR = F_measuredF_standard × 100%

Where:

• F_measured = load required to produce 0.1-in. penetration (lb/in.²)
• F_standard = 1,000 psi at 0.1-in. penetration (standard aggregate value)

Soaked CBR (4-day immersion) is used for pavement design to simulate worst-case seasonal saturation. Unsoaked CBR may be used for check only.

Table 18-2: Typical CBR Values by USCS Soil Classification

USCS Symbol	Soil Description	Typical CBR Range (%)	AASHTO Class	Suitability
GW, GP	Clean gravels	40–80	A-1-a	Excellent
GM, GC	Silty/clayey gravels	20–60	A-1-b	Good–Excellent
SW, SP	Clean sands	10–40	A-1-b, A-3	Good
SM	Silty sands	5–20	A-2-4, A-2-6	Fair–Good
SC	Clayey sands	4–12	A-2-6, A-2-7	Fair
ML	Silt, sandy silt	3–8	A-4	Poor–Fair
CL	Low-plasticity clay	2–6	A-6	Poor
OL	Organic silt/clay	1–4	A-4, A-5	Very Poor
MH	Elastic silt	2–6	A-5	Poor
CH	High-plasticity clay	1–4	A-7-5, A-7-6	Very Poor
OH, Pt	Organic clay/peat	< 2	A-8	Unsuitable

Reference graphics

18.2.3 Modulus of Subgrade Reaction (k)

For rigid pavement (Portland cement concrete) design, the Westergaard modulus of subgrade reaction k (pci — lb/in.³) characterizes the support provided by the subgrade beneath the slab:

Plate load test (ASTM D1196): Apply load via 30-in. diameter bearing plate; record settlement at 0.05-in. increments:

k = pδ    (pci)

Where p = applied pressure (psi) and δ = plate settlement (in.) at 0.1-in. target penetration.

Correlation from CBR (Portland Cement Association):

k (pci) = CBR0.075    (approximate for design purposes)

Correlation from M_r:

k (pci) ≈ M_r19.4    (Ioannides et al. 1990)

Table 18-3: Typical k Values by Soil Type

Soil Type	USCS	CBR (%)	k (pci)	M_r (psi)	Pavement Suitability
Gravel/gravel-sand	GW, GP	40–80	535–1,070	15,000–30,000	Excellent
Sandy gravel	GM, GC	20–50	267–667	8,000–20,000	Good
Sand, gravelly sand	SW, SP	10–40	133–535	5,000–15,000	Fair–Good
Silty sand	SM	5–20	67–267	3,500–8,000	Fair
Clayey sand	SC	4–12	53–160	3,000–5,500	Fair
Silt, sandy silt	ML	3–8	40–107	2,500–4,500	Poor–Fair
Low-plasticity clay	CL	2–6	27–80	1,500–3,500	Poor
High-plasticity clay	CH	1–4	13–53	1,000–2,500	Very Poor

Reference graphics

CBR and modulus ranges for the UIC subgrade classes

UIC links its subgrade quality classes (S0–S3) to ranges of modulus of elasticity and CBR (Figure 18-11). This lets CBR results from yard or access-road work be read against the track-subgrade classification in Chapter 2. Rock subgrades have moduli of the order of 430,000 psi. Profillidis defines CBR as the load producing 0.1 in. of penetration in the sample, divided by the load producing the same penetration in a standard reference material.

He adapts Dormon’s rule from highway engineering for the effect of traffic on subgrade stress: σ₁/σ₂ = (N₂/N₁)^λ, with λ ≈ 0.2.

Figures · click to zoom

After Profillidis (2014), §9.6, §9.8, Fig. 9.5. Full reference under Supplemental Sources.

Estimating subgrade strength by feel

Table 18-8. Guide for estimating subgrade soil strength in the field

Feel	Field indication	SPT (blows)	Shear strength (psf)	CBR
Very soft	Person standing sinks > 3 in.	< 2	< 250	< 0.4
Soft	Person walking sinks 2–3 in.	2–4	250–500	0.4–0.8
Medium	Person walking sinks ~1 in.	4–8	500–1,000	0.8–1.6
Stiff	Pickup ruts 0.47–1 in.	8–15	1,000–2,000	1.6–3.2
Very stiff	Loaded dump truck ruts 1–3 in.	15–30	2,000–4,000	3.2–6.4
Hard	Insignificant ruts under loaded dump truck	> 30	> 4,000	> 6.4

Values as given in the course reference; use for screening, not design.

Figures · click to zoom

After Geotechnology for Railroaders Program, Module 2; §10. Full reference under Supplemental Sources.

Case study: 18.15.1 Chasing a CBR of 7% into competent rock (Look 2023), in §18.15 Case Studies.

For jointed plain concrete pavement (JPCP), the AASHTO 1993 rigid pavement design equation is:

Where:

• D = slab thickness (in.)
• S'_c = concrete modulus of rupture (psi); typically 600–700 psi for 4,000-psi concrete
• C_d = drainage coefficient (0.80–1.20)
• J = load transfer coefficient (3.2 for doweled joints; 4.4 for undoweled)
• E_c = elastic modulus of concrete (psi); E_c = 57,000√f'_c ≈ 3.6 × 10⁶ psi for f'_c = 4,000 psi
• k = effective modulus of subgrade reaction (pci)
• p_t = terminal serviceability (typically 2.5 for heavy industrial)

Reference graphics

18.6.1 Compaction Standards

Subgrade preparation specifications reference ASTM D698 (Standard Proctor) or ASTM D1557 (Modified Proctor):

RC = γ_d,fieldγ_d,max × 100%

Table 18-6: Subgrade Compaction Criteria by Railroad Client

Railroad	Test Method	Min RC (Top 12 in.)	Min RC (Below 12 in.)	Moisture Range	Notes
UPRR	ASTM D1557 (Modified)	95%	92%	−2% to +2% of w_opt	Intermodal yards; Class I requirement
BNSF	ASTM D1557	95%	92%	−2% to +2%	Heavy terminal areas
CSXT	ASTM D698 or D1557	95% (D1557)	92%	−1% to +3%	D1557 for HMA areas
NSR	ASTM D698	95%	90%	−1% to +3%	General yard areas
Amtrak	ASTM D1557	95%	93%	±2%	Station and platform areas
CPKC	ASTM D1557	98%	95%	−2% to +1%	Frost heave mitigation
CN	ASTM D1557	98%	95%	−2% to +1%	Prairie terminal standard
G&W	ASTM D698	95%	90%	−1% to +3%	State DOT or AREMA
AREMA	ASTM D698	95%	90%	±3%	Baseline standard

Reference graphics

Reference graphics · click to zoom

18.6.2 Subgrade Stabilization

When in-situ subgrade does not meet minimum CBR or M_r criteria, stabilization options include:

• Lime stabilization: Add 3–8% hydrated lime (ASTM C977) to reduce PI and increase M_r. PI reduction target: PI_final < 12 for CL, CH soils. M_r improvement: typically 1.5–3× improvement over unstabilized value.
• Cement stabilization: Add 4–8% Portland cement (ASTM C150) to achieve unconfined compressive strength q_u ≥ 300 psi at 7 days (ASTM D1633). This yields a_i = 0.23 per Table 18-4.
• Geosynthetic reinforcement: Biaxial geogrids (ASTM D6637; aperture size 1.0–1.5 in.) placed at subgrade/subbase interface to improve bearing and reduce differential settlement. Aperture Stability Modulus (ASM) ≥ 0.32 N·m/deg preferred for heavy container crane loads.
• Full subgrade replacement: Remove and replace F3/F4 materials to frost depth or 5 ft below finished grade, whichever governs.

Reference graphics

Supplemental figures · LEVI Library

Case study: 18.15.2 Desiccation cracks cut subgrade modulus by about a third (Look 2023), in §18.15 Case Studies.

Intermodal yard overhead gantry cranes (Rubber Tired Gantry — RTG, or Rail Mounted Gantry — RMG) impose wheel loads of 75–150 kips per wheel on crane rail foundation beams. The geotechnical design challenge is controlling differential settlement between adjacent crane rail beam pile caps:

Differential settlement criterion:

For a 50-ft span between crane rail beam pile caps: ΔS_allow = 50 × 12 in. / 500 = 1.2 in. maximum differential. CN and CPKC tighten this to L/750 (0.80 in.) for high-cycle RTG operations.

Crane rail foundation design is addressed in Chapter 11 (bearing capacity) and Chapter 12 (settlement). The subgrade evaluation output for crane rail areas includes:

• Subgrade CBR (minimum 10 for unpaved laydown zones; 15 for crane lane shoulders)
• Settlement profile under crane wheel loads (Boussinesq stress distribution at depth)
• Differential settlement estimate between pile cap locations
• Subgrade variability assessment (coefficient of variation of CBR along crane runway ≤ 25%)

Case study: 18.15.3 Temporary haul road for a one-off heavy load (Look 2023), in §18.15 Case Studies.

Worked Example 18-1: AASHTO 1993 Flexible Pavement Design — BNSF Intermodal Yard, Kansas City

Project: New reach stacker lanes at BNSF Murray Yard expansion, Kansas City, MO. Design life 20 years; target reliability R = 95%.

Given:

• W₁₈ = 8,000,000 ESALs (20-yr design)
• In-situ subgrade: CL clay; CBR = 4% (soaked); M_r = 1,500 × 4 = 6,000 psi
• Z_R = −1.645 (R = 95%); S₀ = 0.45
• ΔPSI = 4.2 − 2.0 = 2.2 (heavy industrial; p_t = 2.0)
• Drainage coefficient m₂ = m₃ = 0.90 (fair; storm drainage provided)

Step 1 — Solve AASHTO equation for SN (iterative):

Trial SN = 6.0 in.:

Trial SN = 4.5 in.:

Interpolation yields SN ≈ 5.0 in. at W₁₈ = 8,000,000.

Step 2 — Layer design:

SN = 3.76 < 5.0 — deficient. Revise:

Step 3 — Subgrade improvement check:

CBR = 4% < BNSF minimum of 8% (Table 18-1). Lime-stabilize top 12 in. of CL subgrade with 5% hydrated lime → target PI_final < 12 → M_r,improved ≈ 9,000 psi → re-solve for SN:

With M_r = 9,000 psi, SN requirement drops to ≈ 4.1 in. Total construction savings from lime stabilization: reduce GSB from 12 in. to 6 in. and base from 10 in. to 8 in., saving approximately 12 in. of aggregate at ~$18/ton = $2.20/ft² savings over the terminal footprint.

Recommendation: 5% lime stabilization of subgrade + 6 in. HMA + 8 in. CAB + 6 in. GSB (SN = 4.41 in. > 4.1 in. required).

Worked Example 18-2: Rigid Pavement — CPKC Intermodal Yard, Winnipeg

Project: New RTG crane runway and reach stacker lanes, CPKC Winnipeg Intermodal Terminal expansion. Frost depth = 8.0 ft design. Design life 25 years.

Given:

• W₁₈ = 12,000,000 ESALs; R = 95%; S₀ = 0.35; ΔPSI = 2.5
• Subgrade: SM (silty sand); CBR = 7% (soaked); k = 7/0.075 = 93 pci
• Frost susceptibility: F3 (high) — full replacement to 8.0 ft with granular material required
• Post-replacement subbase: 8.0 ft of granular subbase (CBR ≥ 40%) → k_effective = 200 pci
• Concrete: f'_c = 4,500 psi → S'_c = 650 psi; E_c = 57,000√4,500 = 3.82 × 10⁶ psi
• J = 3.2 (doweled joints at 15-ft spacing); C_d = 1.0; p_t = 2.5

Step 1 — Solve rigid pavement design equation for D (iterative):

Trial D = 11 in.:

Try D = 10 in. → verify W₁₈ ≈ 8.5M (< 12M target). D = 11 in. governs.

Recommendation: 11-in. JPCP with #5 dowel bars at 12-in. spacing at transverse joints; 15-ft joint spacing; PCC f'_c = 4,500 psi. Subgrade: full replacement of F3 material to 8.0 ft with Type A granular fill (non-frost-susceptible, P-200 ≤ 3%). Geotextile separator at subgrade interface (CPKC TS Geotextile Specification).

Worked Example 18-3: Frost Penetration and Seasonal M_r — CN Prairie Terminal

Project: CN Winnipeg North Transfer Facility, MOB access roads. Winter design required.

Given:

• Subgrade: CL clay; w = 22%; γ_d = 95 pcf; M_r(summer) = 5,500 psi; M_r(thaw) = 0.25 × 5,500 = 1,375 psi
• Design freezing index F_I = 3,500 °F-days (99% confidence)
• Thermal conductivity (frozen): k_f = 1.10 BTU/hr·ft·°F
• Latent heat: λ = 1.08 × 0.22 × 95 = 22.6 BTU/ft³

Step 1 — Frost depth:

Step 2 — Weighted subgrade M_r (harmonic mean, 12-month basis):

Step 3 — Conclusion: Design subgrade M_r = 3,440 psi (seasonal average governs over 5,500 summer value). Required SN increases by approximately 18% relative to a non-frost-corrected design — equivalent to adding 2 in. of HMA or 5 in. of granular base. CN ES requires the frost-depth granular replacement to eliminate M_r thaw reduction at critical subgrade depths.

All CBR values at Modified Proctor 95% RC, soaked 4 days per ASTM D1883. M_r estimated from CBR using Heukelom & Klomp (CBR ≤ 10) and NCHRP 1-28A (CBR > 10).

Figure 18-1: AASHTO 1993 Nomograph — SN vs. W₁₈ for varying M_r values (3,000 to 15,000 psi) at R = 95%, S₀ = 0.45, ΔPSI = 2.2. BNSF Kansas City design point plotted.

Figure 18-2: Pavement cross-section schematic — intermodal yard flexible pavement (lime-stabilized subgrade + GSB + CAB + HMA), annotated with layer thicknesses and material specifications per BNSF EI.

Figure 18-3: JPCP cross-section — CPKC Winnipeg crane runway (11-in. slab + 8-ft granular frost protection + geotextile separator + CL subgrade), with dowel bar detail at transverse joint.

Figure 18-4: CBR vs. M_r correlation chart — both Heukelom & Klomp and NCHRP 1-28A relationships plotted with laboratory data points from Table 18-7.

Figure 18-5: Frost penetration depth Z_f vs. Freezing Index F_I for k_f = 0.75, 1.00, 1.25, 1.50 BTU/hr·ft·°F — curves covering F_I range of 500–4,000 °F-days with CN and CPKC design points highlighted.

Figure 18-6: Seasonal M_r variation profile — 12-month plot of M_r(t) for CN Prairie terminal showing summer, freezing, frozen, and thaw periods with harmonic mean M_r = 3,440 psi annotated.

Figure 18-7: Intermodal yard plan view — test pit and CBR sampling layout showing reach stacker lanes, crane runway, container stacking grid, and areas of F2 vs. F3 frost susceptibility (from Table 18-7 results).

Figure 18-8: Subgrade improvement comparison — untreated CL vs. 5% lime-stabilized CL, showing PI reduction, CBR improvement, and M_r improvement with pavement SN reduction.

Figure 18-9: Container crane RTG wheel load distribution diagram — crane rail beam, pile caps, and subgrade stress bulb at 2:1 distribution showing stress at subgrade surface for 120-kip RTG wheel load.

Figure 18-10: ESAL accumulation curves — BNSF Kansas City terminal projected W₁₈ vs. years of operation for low, medium, and high throughput scenarios, with 20-yr design ESAL = 8M annotated.

Reference graphics

1. Test the subgrade at finished-grade elevation, not natural grade. Many intermodal yard designs involve cut-and-fill grading of 3–8 ft. The pavement subgrade is the constructed fill surface, not the existing ground surface. Obtain CBR and M_r from test pits or undisturbed samples at the design subgrade elevation. Conducting CBR testing only at the surface of existing ground produces non-representative values when significant earthwork is planned.
2. Perform soaked CBR as the design value — never unsoaked. Intermodal yards in humid regions (CSXT Southeast, NSR Mid-Atlantic) experience frequent storm events and elevated seasonal groundwater. Unsoaked CBR values commonly exceed soaked CBR by 3–8× for CL and CH soils. BNSF, UPRR, CSXT, and CN all require soaked CBR per ASTM D1883 as the governing design value.
3. Separate crane rail foundation design from reach stacker lane design. RTG crane rails impose point loads 5–10× greater than reach stacker distributed loads. Design crane rail foundation beams on piles (Chapter 11) independently of the pavement structural number calculation. Intermodal crane runway serviceability is governed by differential settlement (Chapter 12), not pavement structural number.
4. Address frost-susceptible materials at design stage, not during construction. For CPKC and CN projects, frost susceptibility classification (Table 18-5) must be completed during the geotechnical investigation, not deferred to construction. Discovering F4 material at frost depth during grading operations triggers costly change orders and schedule impacts. Budget for complete soil profiling to Z_f + 3 ft at all facility locations above 44°N latitude.
5. Apply lime or cement stabilization conservatively in freeze-thaw zones. Lime-stabilized clay subgrades in climates with frequent freeze-thaw cycling (Winnipeg, Chicago, Minneapolis) may experience swell and strength loss due to ettringite formation if sulfate content exceeds 2,000 ppm (SO₄). Test for sulfates (ASTM D4327) before specifying lime treatment at CPKC and CN prairie sites. Cement or fly-ash alternatives may be preferred where sulfates are elevated.
6. Document subgrade variability along crane runway alignments. Crane runway alignment over variable subgrade produces differential settlement that disrupts crane operations. Minimum investigation for crane runways: CBR at 50-ft intervals along each rail, with test pits to 10-ft depth. Flag locations where CBR coefficient of variation exceeds 25% for additional investigation or localized improvement.
7. Include a pavement management monitoring program in deliverables. All Class I railroad clients (UPRR, BNSF, CSXT, CN, CPKC) maintain facility pavement management systems. The geotechnical GDR should include a subgrade baseline data table suitable for direct import into the client's PAVER or similar pavement management system, with unique asset IDs, GPS coordinates, subgrade M_r, design ESAL, and year of construction for each pavement section.

Reference graphics

Reference graphics · click to zoom

Project management · Bull (2015)

Counting Closures in a Life-Cycle Cost Comparison: The I-710 Case

In Bull’s volume, Lee, Kim and Harvey (Chapter 4) check the pavement type chosen for Caltrans’ I-710 Long Beach rehabilitation. They follow the Caltrans life-cycle cost procedure: a 60-year analysis period, a 4% real discount rate and no salvage value. Three alternatives were compared: innovative long-life asphalt (the design actually built), standard asphalt, and long-life concrete. Two tools were used. CA4PRS estimated the construction schedule, the number of closures, and the work-zone user delay for the initial work and for every future maintenance and rehabilitation event. RealCost then discounted the agency and user costs.

Agency cost started from unit prices × quantities, taken from the contractor bid database. Allowances for traffic handling, drainage, storm-water protection, minor items, mobilization, supplemental work and a 20% contingency were then applied as cumulative multipliers. Together these multiplied the bare pavement cost by 1.80.

The closure count proved central. The built design needed 10 fifty-five-hour weekend closures at first and about 296 seven-hour night closures over 60 years. The concrete option needed 30 weekend closures at first, a further 18 at year 50, and about 148 nights.

Table 18-PM1. Table: 60-year present values (after Lee, Kim and Harvey in Bull 2015, Table 4.3)

Alternative	Agency cost	User cost	Total
Innovative long-life AC	$33.2M	$5.8M	$39.0M
Standard AC	$41.1M	$12.1M	$53.2M
Long-life PCC	$50.4M	$9.8M	$60.2M

The standard asphalt option had the lowest initial cost but the highest life-cycle user cost, about twice that of the design that was built.

For railroad work: for intermodal yard pavements, crane-rail runways and grade-crossing approaches, compare subgrade treatment and pavement options the way Caltrans did. For each alternative:

• Lay out its maintenance and rehabilitation sequence over the full period.
• Count the yard-track, lane or crane outages each event needs.
• Price those outages as a user cost, using lost lifts, terminal dwell or train delay.

The same accounting applies to track itself: count the track windows each subgrade option needs over its life, not only the windows for first construction. State the markups explicitly, as the cumulative multiplier table does, so that a reviewer can see how much of the estimate is contingency.

Source: Bull, J. W. (ed.) (2015). Life Cycle Costing for the Analysis, Management and Maintenance of Civil Engineering Infrastructure. Dunbeath, Scotland: Whittles Publishing, pp. 91–108. Summarized and paraphrased; figures reproduced for internal reference. Full reference under Supplemental Sources.

Question 10-1: A CSXT intermodal yard in Jacksonville, FL has subgrade CL with soaked CBR = 5%. Estimate M_r using the Heukelom & Klomp equation. What AASHTO 1993 SN is required for W₁₈ = 10,000,000 ESALs, R = 95%, S₀ = 0.45, ΔPSI = 2.2?

Answer 10-1:

• M_r = 1,500 × CBR = 1,500 × 5 = 7,500 psi
• Using AASHTO equation, iterative solution for SN:

Trial SN = 5.0: log W₁₈ = (−1.645)(0.45) + 9.36 log(6.0) − 0.20 + [log(2.2/2.7)] / [0.40 + 1094/(6.0)^5.19] + 2.32 log(7,500) − 8.07

(6.0)^5.19 = e^(5.19 × ln 6.0) = e^(5.19 × 1.792) = e^9.30 = 10,938

= −0.740 + 9.36(0.778) − 0.20 + (−0.089)/(0.40 + 0.100) + 2.32(3.875) − 8.07

= −0.740 + 7.282 − 0.20 − 0.178 + 8.990 − 8.07 = 7.084

W₁₈ = 10^7.084 = 12,140,000 (slightly high; SN = 4.8 in. governs) → SN ≈ 4.8 in.

Layer design: 5 in. HMA (SN_1 = 2.20) + 8 in. CAB (SN_2 = 0.14 × 8 × 0.90 = 1.01) + 12 in. GSB (SN_3 = 0.11 × 12 × 0.90 = 1.19) → SN_total = 4.40 in.

Revise: 6 in. HMA + 8 in. CAB + 10 in. GSB → SN = 2.64 + 1.01 + 0.99 = 4.64 in. ≈ required. Use 6–8–10 section; add lime stabilization to raise CBR from 5% to 8% target for CSXT.

Question 10-2: A CPKC Prairie facility in Regina, SK has freezing index F_I = 3,200 °F-days. Subgrade is ML (elastic silt) with w = 18%, γ_d = 102 pcf. Thermal conductivity k_f = 1.05 BTU/hr·ft·°F. Compute frost depth and frost susceptibility classification.

Answer 10-2:

• λ = 1.08 × 0.18 × 102 = 19.8 BTU/ft³
• Z_f = √(48 × 1.05 × 3,200 / 19.8) = √(161,280/19.8) = √8,145 = 90.3 in. = 7.5 ft
• ML (elastic silt): typical heave rate 0.010–0.025 in./day → Frost Susceptibility Group F4 (Very High)
• CPKC TS requires full replacement of ML subgrade to 7.5-ft depth with non-frost-susceptible granular material (P-200 ≤ 3%, CBR ≥ 40%). Subgrade drainage blanket (12-in. granular; geotextile-wrapped) required at subgrade interface.

Question 10-3: A reach stacker at an NSR Atlanta terminal has a tandem drive axle load of 110,000 lb. How many ESALs does one pass (loaded outbound + empty return) represent, using the 4th-power rule? If the terminal handles 300 container moves per day, what is the annual ESAL count?

Answer 10-3:

• For tandem axle load P = 110,000 lb: standard tandem = 32,000 lb ESAL_factor = (110,000/32,000)^4.0 = (3.4375)^4.0 = 139.7 ESALs per loaded pass
• Empty return: assume 60% of loaded weight = 66,000 lb → ESAL = (66,000/32,000)^4.0 = (2.0625)^4.0 = 18.1 ESALs
• ESALs per container move = 139.7 (loaded) + 18.1 (empty) = 157.8 ESALs/move
• Annual: 300 moves/day × 157.8 ESAL/move × 365 days = 17,282,700 ESAL/yr
• This significantly exceeds the 8M NSR terminal target (Table 18-1), requiring either a revised traffic count, higher reliability design, or rigid (JPCP) pavement to extend pavement life.

Question 10-4: A UPRR intermodal yard subgrade has M_r = 4,500 psi from laboratory testing. What is the estimated k modulus (pci) for rigid pavement design? What concrete slab thickness D is required if UPRR targets W₁₈ = 15M ESALs, R = 95%, S₀ = 0.35, S'_c = 650 psi, J = 3.2, C_d = 1.0, p_t = 2.5?

Answer 10-4:

• k = M_r/19.4 = 4,500/19.4 = 232 pci
• AASHTO rigid design — trial D = 11 in.: E_c = 57,000√4,500 = 3.82 × 10⁶; E_c/k = 3,820,000/232 = 16,466; (E_c/k)^0.25 = 11.33 D^0.75 = 11^0.75 = 6.64; 18.42/11.33 = 1.626; D^0.75 − 18.42/(E_c/k)^0.25 = 6.64 − 1.626 = 5.014 Numerator: 650 × 1.0 × (6.64 − 1.132) = 650 × 5.508 = 3,580 Denominator: 215.63 × 3.2 × 5.014 = 3,458 log ratio = log(3,580/3,458) = log(1.035) = 0.015

(12)^8.46 term → 1+1.624×10⁷/(12)^8.46 ≈ 1.012 (from previous computation) log W₁₈ = −(1.645)(0.35) + 7.35 log(12) − 0.06 + (−0.079/1.012) + (4.22 − 0.80)(0.015) = −0.576 + 7.930 − 0.06 − 0.078 + 0.051 = 7.267 → W₁₈ = 18.5M D = 11 in. supports 18.5M > 15M → D = 10.5 in. (govern; specify 11 in. for constructability) UPRR recommended section: 11-in. JPCP with 6-in. CAB base, #5 dowel bars at 15-ft joints.

Question 10-5: CN geotechnical data shows a CH subgrade (CBR = 2%) beneath a proposed Thunder Bay, ON intermodal yard. AASHTO Group Index (GI) must be reported. Given: LL = 62, PL = 22, P-200 = 88%. Is this soil suitable as a subgrade material? What treatment is recommended?

Answer 10-5:

• PI = LL − PL = 62 − 22 = 40
• F = P-200 − 35 = 88 − 35 = 53 (truncate to 40 maximum for GI formula)
• F' = P-200 − 15 = 88 − 15 = 73 (truncate to 40 maximum)

• AASHTO Class A-7-6; GI = 24 → Very poor subgrade; unsuitable as pavement subgrade.
• M_r = 1,500 × 2 = 3,000 psi (minimum bound); frost susceptibility F4 (Very High) at Thunder Bay.
• CN ES required treatment: (1) Remove and replace CH to frost depth (Z_f ≈ 6.5 ft, Table 18-5); (2) Install 6-in. granular drainage blanket at subgrade interface; (3) Place non-frost-susceptible granular fill to finished subgrade elevation; (4) Alternatively, consider deep soil mixing (Chapter 17) if removal depth exceeds 8 ft and crane rail differential settlement governs.

End of Chapter 18 — Pavement Subgrade Evaluation

Field case studies by Burt G. Look, from Earthworks: Theory to Practice – Design and Construction (CRC Press/Balkema, 2023), chosen for this chapter’s topics. Each is summarized from the book with its key figures and a closing note on railroad application added by the authors of this manual. Credit for the case material belongs to Look; see item S10 under Supplemental Sources.

Case Study 18.15.1 — Chasing a CBR of 7% into competent rock

Case study · Look (2023) §6.11

Related section: 18.2 Subgrade Strength Testing

Urban road works at Kelvin Grove, Brisbane (2002) required a soaked subgrade CBR of 7% over residual soil and near-surface phyllite. The weathered rock at subgrade level degraded to soil after rain. Samples were taken from the degraded surface with its clay seams, and the laboratory removed the “stone” sizes. The CBR came back at 5% or less. Two further rounds of excavation, each needing a rock-breaker attachment to reach the next sample depth, also failed, ending 4.9 ft below design level, with about AU$100,000 of claims for “unsuitable subgrade” and delays in a busy urban area.

Look’s review found that only the top 4 in had degraded. Below it was competent rock (it needed a rock breaker) with a CBR well above 10%, and once the pavement covered it no further degradation would occur. The tests themselves were unrepresentative because the stone fraction was discarded, so the low CBR could be disregarded.

For railroad work: the same cycle can play out under a new track or yard pavement in weathered shale or phyllite. Protect the formation from rain, trim the softened skin, and check whether the sample is representative before ordering more over-excavation.

Source: Look, B. G. (2023). Earthworks: Theory to Practice – Design and Construction. CRC Press/Balkema. Case study §6.11, pp. 192–194. Summarized and paraphrased; figures reproduced for internal reference. Full reference under Supplemental Sources.

Case Study 18.15.2 — Desiccation cracks cut subgrade modulus by about a third

Case study · Look (2023) §10.16

Related section: 18.6 Subgrade Preparation and Compaction Criteria

An expansive clay subgrade passed its density tests (above 95% standard MDD) and was then left uncovered for several weeks of wind and heat. Desiccation cracks appeared, and the question was whether the design modulus of 7,200 psi (CBR about 5%) still held. Plate load tests on cracked and uncracked areas, using the site roller as reaction, showed the cracked areas reached about 60% of the uncracked modulus on the first load cycle and 70% on the second, which is used for fill. The target was EV2 above 7,200 psi at formation level and above 4,350 psi below it. A one-third loss of strength or modulus is typical for fissured or desiccated clays in design.

Table 18-LK1. Plate load tests on uncracked and cracked compacted clay (after Look 2023, Table 10.18)

Load cycle	Uncracked E (psi)	Cracked E (psi)	Ratio
First	4,600	2,760	~60%
Second	11,600	8,100	70%

For railroad work: a finished clay subgrade awaiting subballast should be covered or kept moist. If it cracks, re-test it before placing subballast; a passing density test from before the drying no longer applies.

Source: Look, B. G. (2023). Earthworks: Theory to Practice – Design and Construction. CRC Press/Balkema. Case study §10.16, pp. 366–368. Summarized and paraphrased; figures reproduced for internal reference. Full reference under Supplemental Sources.

Case Study 18.15.3 — Temporary haul road for a one-off heavy load

Case study · Look (2023) §11.7

Related section: 18.7 Intermodal Yard Crane Rail Foundation Subgrade

Wind-farm and substation access roads may carry a single heavy delivery, where a bogged transporter can cost over a million dollars. A 22-ton or 44-ton axle exerts about 92 or 116 psi at the running surface. These are wheeled loads, so the BR470 working-platform method for tracked plant does not apply, and neither does ordinary road pavement design; the Giroud and Han (2004) method is typically used. Roads and hardstands typically need 5,200 and 6,300 psf allowable bearing, more on grades above about 5%.

Soaked tests and drainage are usually left out of a one-off design for cost reasons, so rain is the main risk: Look treats more than about 0.8 in of rain as high risk when an unsoaked CBR was used. At this site the proof roll failed the day before the move, after 2.1 in. and 1.1 in. of rain two and three days earlier. No passing proof roll was seen, but the equipment went on anyway, with significant consequences. Deep ruts under one wheel and none under the other showed non-uniform construction or support.

Recommended field data: DCP, vane shear, light falling-weight deflectometer and plate load tests, with plate tests kept to hardstands and critical spots. The maximum plate pressure reached is not the allowable bearing capacity. In one example a 1,250 psf design subgrade strength, with influence extending 24–31 in, gave calculated ruts above 2 in.

Table 18-LK2. Ground contact pressure by axle load (after Look 2023, Table 11.5)

Axle load (tons)	Wheel load (tons)	Contact area (in.²)	Contact pressure
13	6.6	167	80 psi
22	17	240	92 psi
44	22	380	116 psi

For railroad work: the same logic applies to crane pads and temporary access for bridge erection, transformer delivery to traction substations and intermodal heavy lifts. Hold the move on a failed proof roll after rain.

Source: Look, B. G. (2023). Earthworks: Theory to Practice – Design and Construction. CRC Press/Balkema. Case study §11.7, pp. 390–392. Summarized and paraphrased; figures reproduced for internal reference. Full reference under Supplemental Sources.

Authors: Daniel P. Colwell and Mario Ruel. This chapter was written and compiled by Daniel P. Colwell and Mario Ruel for Railway Ground Hazard Risk Management.

Technical content is drawn from the published sources below, and credit for that material belongs to their authors and publishers. Tables, equations and figures taken from a specific source are also cited where they appear.

Primary sources

• USACE EM 1110-1-1804, Geotechnical Investigations (2001)
• AASHTO Guide for Design of Pavement Structures (1993)
• AASHTO M 145
• ASTM D1883 (CBR); ASTM D4318 (Atterberg limits)
• NCHRP Project 1-37A (resilient modulus)
• AREMA MRE Chapter 1
• FRA Track Safety Standards, 49 CFR Part 213

Supplemental sources (LEVI Library)

• Huang, Y. H., Rose, J. G. and Khoury, C. J. (c. 1987). “Thickness Design for Hot-Mix Asphalt Railroad Trackbeds.” Proc. Association of Asphalt Paving Technologists, pp. 427–453 (scanned; text recovered by OCR)
• Limburg, A. and Nielsen, D. (2019). “Fast-Tracked Pavement Maintenance Using Concrete Rubblization.” AREMA 2019 Annual Conference Proceedings
• SHRP2 R02 / GeoTechTools (2012). Geosynthetic Separation in Pavement Systems — fact sheet, design guidance and photographs
• SHRP2 R02 / GeoTechTools (2012). Geocell Confinement in Pavement Systems — fact sheet, design guidance and photographs
• U.S. Army Corps of Engineers (2010). PCASE 2.09 User Manual. Transportation Systems Center / ERDC
• Mauduit, C., Livet, J. and Piau, J-M. (2007). “Frost Design Method for Roadways and Railways – State of the Art in France.” In Gomes Correia, A., Momoya, Y. and Tatsuoka, F. (eds.), Design and Construction of Pavements and Rail Tracks. Taylor & Francis/Balkema
• Look, B. G. (2023). Earthworks: Theory to Practice – Design and Construction. CRC Press/Balkema. Case studies §6.11, §10.16, §11.7

File locations for these documents are listed under Supplemental Sources — LEVI Library.

Project management sources

• Bull, J. W. (ed.) (2015). Life Cycle Costing for the Analysis, Management and Maintenance of Civil Engineering Infrastructure. Dunbeath, Scotland: Whittles Publishing