International Concrete Crosstie & Fastening System Symposium Investigation of the Mechanics of Rail Seat Deterioration (RSD) and Methods to Improve the Abrasion Resistance of Concrete Sleepers International Concrete Crosstie & Fastening System Symposium Urbana, IL 6-8 June 2012 Part of larger tie and fastener research program sponsored by FRA Ryan G. Kernes, Amogh A. Shurpali, J. Riley Edwards, Marcus S. Dersch, David A. Lange, and Christopher P.L. Barkan

Outline Objectives Rail seat deterioration (RSD) background Large scale abrasion test Evaluation of frictional properties Small scale abrasion resistance test Results Conclusions and future work

Objectives Understand the mechanics of the most critical failure modes Propose methods of mitigating the critical failure modes Investigate parameters that affect abrasion mechanism Quantify abrasion resistance of various concrete mix designs, curing conditions, and surface treatments Characterize frictional forces between rail seat and rail pad Through industry surveys and focused discussions with industry experts Understand mechanics by experimentation or modeling

Rail Seat Deterioration (RSD) Degradation of concrete material under rail and pad Increases maintenance costs Shortens service life of the concrete crosstie Leads to track geometry defects Feasible mechanisms: Abrasion Crushing Freeze-thaw damage Hydro-abrasive erosion Hydraulic pressure cracking Track geometry defects could result in an unstable rail condition

Abrasion Mechanism of RSD Abrasion is a progressive failure mechanism that occurs when: Frictional forces act between two surfaces in contact Relative movement occurs between the surfaces Harder surface cuts or ploughs into softer surface Progression of abrasion at the rail seat Cyclic motion of rail base induces shear forces Shear forces overcome static friction Pad slips relative to concrete Strain is imparted on concrete matrix Abrasion involves 3-body wear: two interacting surfaces (rail pad and rail seat) and abrasive slurry (water and fines) Not an intrinsic property but a system property. Therefore, pads, environmental characteristics, must be investigated. In order to investigate this complex phenomena, we have developed a laboratory setup at the University of Illinois.

Large Scale Abrasion Resistance Test Experimental Setup Vertical Actuator Abrasion Pad Horizontal Actuator Horizontally mounted actuator produces displacements of a pad relative to a fixed concrete specimen A static normal force is applied with a vertically mounted actuator Objectives: Isolate abrasion mechanism Maintain representative contact mechanics Understand effect of variables on abrasion rate Load magnitude, displacement, load rate, water, sand Establish most abrasive parameters to evaluate rail seat materials and surface treatments Test Setup: Nominal pressure: up to 900 psi Displacements: 1/16” – 1/8” 6” x 6” concrete specimens, f’(c) > 7000 psi 4” x 3” stock pad materials, Nylon 6/6 (unreinforced), Polyurethane (95 Shore A) Contaminates: laboratory sand, water Concrete Specimen

Large Scale Abrasion Resistance Test Results Consistently able to cause deterioration of concrete due to abrasion Concrete deterioration initiates near pad edges and propagates inward Heat build up in pad materials at local contact points leads to softening and adhesion to concrete surface Difficulty in correlating severity of abrasion to input variables Contact angle and pressure distribution Heterogeneity of concrete surface Due to the difficulties in correlating abrasion severity, we decided to shift gears with the testing protocol. Due to physical setup of the testing equipment where we can constantly monitor both lateral and vertical loads throughout the test, we are able to gain information related to the friction at the contact interface. So why is friction important?

Experimental Evaluation of Friction Magnitude of static frictional force is directly related to the normal force between the two bodies by the coefficient of friction (μ) Experimental frictional coefficient measured during test calculated by: μ = |𝑭|/𝑷 where F is the force required to initiate lateral sliding under vertical normal load P 400 loading cycles to simulate single unit train pass 5,000 pound normal vertical load (420 psi), 1/8” lateral displacement, 3 cycles/second, no abrasive fines or water added Because friction is not understood, we devised an experimental protocol to measure the friction at the pad – rail seat contact interface. Pads allowed to cool to original temperature in between tests.

Mean Coefficient of Friction per Loading Cycle Polyurethane Each line on graph represents an average of 4 repetitions of a single pad on a concrete specimen. Overall trends Diff in Poly vs Nylon 6/6 based, changes based on modulus and thermal effects Nylon 6/6 experimental modulus is about 44 KSI Poly experimental modulus is about 30 KSI Overall trend of decreasing COF due to contact films Nylon 6/6

Effect of Environmental Conditions on Nylon 6/6 Average of 16 trials

Effect of Environmental Conditions on Concrete Abrasion Polished surface with no sand Abrasion initiated with sand

Effect of Increasing Normal Load on Nylon 6/6 3 kips 5 kips Average of 8 trials at 3 and 10 kips 5 kip data is average of 44 trials 10 kips

Small Scale Abrasion Resistance Test (SSART): Test Setup In general, similar to other standard abrasion tests Consists of powered rotating steel wheel with 3 lapping rings Lapping rings permitted to rotate about their own axis Vertical load applied using the dead weights Abrasive sand and water dispensed during testing Lap 12 in; Rings 6 in. by 1.5 inch (depth) ; dead weight 4.5 lb Hanging words (more than one) Ottawa sand standard for testing

Test Protocol Each test can evaluate 3 specimens Multiple tests are run to evaluate more than 3 specimens Specimen dimensions: 4 inch (diameter),1 inch (thickness) Duration: 120 minutes Wear depth measurements taken every 20 minutes Speed: 60 revolutions per minute Abrasive fine: Ottawa 20-30 sand Before 20-30: Passing through 841 microns retained on 596 microns Average wear depth: 2.5-6.0 mm/test Average weight loss: 75-100 gm/test. After

Test Variables Admixtures Surface Treatment Silica fume: 5%,10% Fly ash: 15%, 30% Curing Condition Moist Submerged Oven dry Air Surface Treatment UV epoxy Polyurethane Grinding Fiber Reinforced Concrete (FRC) Steel Lapping machine not a representative test, that’s the reason for large scale which is…

Effect of Mineral Admixtures Compressive strength 1-2 specimens ART: 3 ASTM 231 Pressure method Generated results that are in agreement with the prevalent literature related to fly ash and silica fume. This gives us confidence moving forward in evaluating other mixes that have not been tested before.

Effect of Fiber Reinforcement Compressive strength 1-2 specimens ART: 3 ASTM 231 Pressure method

Effect of Curing Conditions

Comparison of Techniques to Increase Abrasion Resistance

Additional Tests Conducted Objective: to obtain specimens batched by industry concrete crosstie suppliers Silica fume and epoxy coated specimens tested Saw-cut surface of control specimens considered analogous to ground surface and tested Saw cut (Sustersic et al 1991) Grinding Epoxy coating

Effect of Surface Treatments

Conclusions Large scale testing: Confirmation of abrasion as a feasible RSD mechanism Frictional coefficient is affected by temperature, water, sand, normal force SSART: Successfully compared 13 approaches to improving abrasion resistance of rail seat through material improvements Improve abrasion resistance of concrete with: Optimal amounts of fly ash Proper curing condition Addition of steel fibers Grinding cement paste

Future Work Determining the optimal frictional properties at each interface of multi-layer rail pads (top, bottom, and between layers) could: Reduce movement at critical interfaces Influence load path and location of slip Delay the onset of abrasive wear and extend rail seat life Extend to other fastening system components SSART: Perform image analysis to characterize the role of coarse aggregate in abrasion resistance Study the effect of air entrainment and quality of aggregates on abrasion resistance of rail seat Use of Statistical Analysis Software (SAS) to model abrasive wear of concrete specimens Optimize concrete mix design and surface treatments to mitigate abrasion

Acknowledgements Funding for this research has been provided by Association of American Railroads Technology Scanning Program NEXTRANS Region V Transportation Center Eisenhower Graduate Fellowship For providing direction, advice, and resources: Amsted Rail - Amsted RPS: Jose Mediavilla, Dave Bowman, Brent Wilson, BNSF Railway: John Bosshart, Tom Brueske, Hank Lees AREMA Committee 30: Winfred Boesterling, Pelle Duong, Kevin Hicks, Tim Johns, Steve Mattson, Jim Parsley, Michael Steidl, Fabian Weber, John Zeman TTCI: Dave Davis, Richard Reiff Pandrol Track Systems: Bob Coats, Scott Tripple UIUC: Tim Prunkard, Mauricio Gutierrez, Don Marrow, Darold Marrow For assisting with the research and lab work: Josh Brickman, Ryan Feeney, Kris Gustafson, Steven Jastrzebski, Andrew Kimmle, Calvin Nutt, Chris Rapp, Amogh Shurpali, Emily Van Dam, Michael Wnek

Questions Ryan Kernes Research Engineer Rail Transportation and Engineering Center - RailTEC email: rkernes2@illinois.edu Amogh A. Shurpali Graduate Research Assistant Rail Transportation and Engineering Center - RailTEC email: ashurpa2@illinois.edu