Transcription

IDENTIFYING FINE AGGREGATES PRONE TO POLISHING IN PCC PAVEMENTSDAVID W. FOWLER & MARC M. RACHEDDepartment of Civil, Environmental and Architectural Engineering, University of Texas atAustin, United [email protected]; marc [email protected] polishing in portland cement concrete (PCC) pavements leads to higherincidences of skid-related accidents on highways. This type of failure is often associatedwith the usage of softer fine aggregate such as limestone sands. To identify polishresistance aggregates, state agencies like TxDOT have adopted tests such as the acidinsoluble residue test (AIR). Since calcium carbonate is soluble in acid, no carbonate sandpasses the AIR test which has a minimum limit of 60% in Texas. This paper describesresearch that was done to evaluate the polish resistance of aggregates using a laboratoryconcrete performance test. Concrete slabs made with different fine aggregates wereevaluated for skid using a circular track meter (CTM), a dynamic friction tester (DFT), anda three-wheel-polishing device (TWPD). To ensure that the values obtained at thelaboratory related to field performance, test sections constructed with 100% limestonesand and blended sands were also evaluated. Results show that some of the aggregatesthat failed the AIR test performed as well as some of the siliceous fine aggregates thatpassed the AIR test. Other aggregate tests such as the micro-Deval have shown to relatemore closely to the concrete performance tests performed under laboratory conditions.1. INTRODUCTIONMany state agencies have set limits on the usage of carbonate sands. In Texas, thecurrent limits are determined by the acid insoluble test residue (AIR) test that has aminimum of 60%. Under current specifications, the maximum quantity of carbonate sandthat can be used in a PCC pavement is less than 40% of the total sand volume since thecarbonate sands generally have an AIR value of less than 10%. Sources of quality naturalsands have begun depleting in some metropolitan areas where the need for concrete ishigh. In such areas the concrete industry has the option to either ship natural sands fromoutside sources or use local sources of manufactured fine aggregates (MFA). Shippingaggregates from outside sources adds to the cost of concrete, and it is important to findmethods to maximize the use of local materials.Several problems arise from using MFAs in PCC pavements: workability, finishability, andskid resistance. These problems exist because of the mineralogy, shape, and grading ofMFAs. While the workability and finishability of concrete can be improved by betterproportioning mixtures, skid resistance is mainly dependent on the mineralogy of the sand.A decrease in skid resistance leads to higher incidences of skid-related accidents onhighways. Softer sands, e.g. carbonate sands, are known to polish when used in PCCpavements and thus provide less long-term skid resistance when compared to hardersiliceous sands. No recent research has been done to evaluate skid resistance of PCCmade with carbonate sands, and thus it is not clear whether or not current specificationsadopted by state agencies accurately reflect the performance of those sands in the field.SURF1-Rached1

This paper presents the results of a research project that was done to evaluate the skidresistance of concrete made with different sands and blends of sands. The goal of thisproject was to find a method to better evaluate the polish resistance of fine aggregatesused in PCC pavements. The research included both field and laboratory testing. Thelaboratory and field skid performance of concrete were evaluated using the Circular TrackMeter (CTM) and Dynamic Friction Tester (DFT). The results obtained from the concretelaboratory testing were then compared to aggregate tests which included the acidinsoluble residue test (AIR) and the micro-Deval abrasion test for fine aggregates.2. BACKGROUNDThe mineralogy of coarse aggregate is vital for obtaining good skid performance in asphaltconcrete. In PCC, however, the mineralogy of the fine aggregate is more important forobtaining good friction. The coarse aggregate only becomes an influencing factor in caseswhere the top surface of the pavement has been severely abraded or when coarseaggregate is intentionally exposed. Folliard and Smith (2003) identified fine aggregatemineralogy and hardness as important factors for obtaining good surface friction after thetexture of a pavement is abraded [1]. Since it is difficult to directly measure the resistanceof fine aggregate to polishing, other indicator tests have been used [1]. The most widelyused test is the acid insoluble residue test (AIR). The test assesses the presence ofnoncarbonated material in the fine aggregate; materials that have high carbonate contentyield low residue because they dissolve in acid, while materials with low carbonate contentyield a high residue. It is believed that the presence of acid insoluble material in the sandfraction generally improves skid resistance [1]. In PCC pavements, the fine aggregatesexposed on the surface constitute the micro-texture (wavelength 0.5 mm, amplitude 1to 500 μm). Micro-texture is important to maintain adequate friction in dry-weatherconditions and wet-weather conditions when speeds are less than 45 mph (72 km/h) [2].Many states have either banned the usage of carbonate fine aggregates in PCCpavements or have required blending those aggregates with harder aggregates to meetcertain limits. In 1958, the need for skid resistant pavements was recognized by the FirstInternational Skid Prevention Conference [3]. After this conference, state agencies starteddeveloping equipment to test skid both in the laboratory and in the field [3]. In 1958, Shupeand Lounsbury showed a correlation between calcium carbonate content of aggregatesand skidding susceptibility [3]. Gray and Renninger (1965) recognized the contribution ofsiliceous sand particles in skid resistance and pioneered the acid insoluble residue test toanalyze the amount of siliceous materials in the aggregates [3]. Balmer and Colley (1966)correlated results of a laboratory concrete skid performance test to the acid insolubleresidue of the aggregates tested. They concluded that 25% siliceous fine aggregatecontent was satisfactory for skid performance with most aggregates. Most specificationsbase their limits on the study done by Balmer and Colley. Some specifications require aminimum of 25% siliceous sand content in pavement concrete, while other specificationshave set limits based on acid insoluble residue (AIR) values. The Texas Department ofTransportation (TxDOT) originally required fine aggregates to meet an AIR limit of 28%,which would have required about 25% siliceous sand content and excluded the usage of100% carbonate sands. After skid problems were reported, the AIR limit was raised to60%. Under the 60% AIR specifications, the maximum amount of carbonate sand that canbe used in a PCC pavement is less than 40% of the total sand volume. The adoption of the60% AIR limit by the TxDOT has affected districts like the Dallas and Ft Worth districts thathave limited local sources of natural siliceous sands and have to haul natural sands fromdistant pits and blend them with their local sources of manufactured carbonate sands tomeet the 60% AIR limit.SURF1-Rached2

Studies done after 1966 had similar conclusions as the study done Balmer and Colley.Renninger and Nichols (1977) found good correlation between skid resistance (asdetermined by the British Pendulum Tester) and acid insoluble residue [4]. As part of astudy that evaluated micro-texture and macro-texture on PCC pavements around theUnited States, Hall and Smith (2009) found that tougher, more durable aggregates retainhigher friction values. They found that the usage of limestone in Kansas and Illinoisresulted in greater rates of micro-texture deterioration compared to the usage of high silicagranite aggregates in Minnesota [2].The Locked-Wheel Skid Trailer (ASTM E 274) is the most common method used toevaluate skid resistance on pavements in the United States. The method consists ofmeasuring the locked-wheel friction (100% slip condition) of a trailer towed behind a truckat a speed of 40 mph (64 km/h) or 50 mph (80 km/h). The trailer administers a water sprayto the pavement in front of the tire to simulate wet conditions. The resulting friction forceacting between the test tire and the pavement surface is used to determine the skidresistance which is reported as a skid number (SN). Higher SN values signify higher skidresistance. A smooth tire (ASTM E 524) or a ribbed tire (ASTM E 501) can be used on theskid trailer. Research has shown that ribbed tires are only capable of evaluating the effectof micro-texture on friction, while smooth tires can measure the contribution of microtexture as well as macro-texture [5, 6]. Some state agencies have trigger skid values thatthey use as means of initiating some sort of rehabilitation treatment; these values differfrom state to state. The most common trigger values reported are SN 35 or 30 for ribbedtires, and SN 20 for smooth tires [6]. It is believed that SN values below those limits canresult in an increase in skid related accidents on roadways.The Locked-Wheel Skid Trailer can only be used in the field, for this reason other devicessuch as the Circular Track Meter (CTM) and the Dynamic Friction Tester (DFT) have beendeveloped to evaluate texture and friction in the laboratory as well as in the field (Figure 1).The Dynamic Friction Tester (DFT) is an apparatus that measures the friction-speedrelationship on a pavement surface for speeds ranging from 0 to 80 km/h (micro-texture).The DFT measures the torque needed to stop three small spring-loaded standard rubberpads rotating in a circular path. The torque measured is then converted to a friction value.Water is also introduced during testing to simulate wet conditions. The Circular TrackMeter (CTM) is a device that utilizes a displacement sensor that is mounted on an arm thatrotates in a circular path and measures the mean profile depth (MPD) of a pavement(macro-texture). The CTM is a device that can be used in the field and laboratory toevaluate macro-texture.Figure 1 – The CTM (left) and the DFT (right)Values obtained from the DFT and CTM can be used to compute an equivalent skidnumber (SN). The correlation between different texture and friction devices wasSURF1-Rached3

established by the Permanent International Association of Road Congresses (PIARC) in1992 [7]. PIARC developed the International Friction Index (IFI), which is an index forcomparing and harmonizing friction measurements with different equipment to a commoncalibrated index. For example, to compute the equivalent skid number (SN) measured by alocked-wheel skid trailer at 40 mph (64 km/h) using a ribbed tire, the following formulascan be used:(eq. 1)(eq. 2)(eq. 3)Where F60 and Sp are the IFI constants, DFT20 is the coefficient of friction measured bythe DFT at 20 km/h, and the MPD is the mean profile depth value measured by the CTM[8].3. TESTING PROGRAM3.1. Materials and Mixture ProportionsFifteen sources and six blends of fine aggregates were evaluated in the laboratory. Thelithology of the sands used in this research was determined by a petrographer. Nine of thesands tested were siliceous with AIR values higher than 60%, while the other six sandswere manufactured sands; these included four limestone, one dolomite, and one slate.The reason more siliceous sands were tested was to evaluate how the difference in theacid insoluble residue value affected skid performance. The six blends of sands testedincluded combinations of siliceous and limestone sands that were blended to meet an AIRvalue of 20%, 40%, and 60%.Two coarse aggregates were used for the laboratory testing. Both were ASTM C 33 grade57 limestone coarse aggregates obtained from two adjacent aggregate pits. The reasontwo sources were used was to include materials from two different producers. Since thecoarse aggregate was not exposed in this study, the source of the coarse aggregate hadminimal effect on the results. Since there were no indications in the literature that themixture proportions of concrete (water-to-cementitious ratio, sand-to-aggregate ratio, orpaste content) influenced skid resistance, all sands tested in concrete were evaluatedusing one standard mixture (Table 1).Table 1 – Mixture Proportions used for Evaluating Fine AggregatesMaterials (Volume %)Cementitious Water Fine Aggregate (SSD) Coarse Aggregate (SSD)10.714.2027.146.0The cement used for all mixtures was an ASTM C 150 Type I/II cement. Since the samemixture proportions were used to make concrete specimen using the different sands, amid-range water reducing admixture was used. This type of admixture is not common forslipform paving mixtures but was used to facilitate casting the concrete specimens. Usinga mid-range instead of regular ASTM C 494 Type A water-reducer might affect theSURF1-Rached4

workability and finishability of the concrete, but was deemed to have no effect on skidresistance.3.2. Test Methods and ProceduresThe acid insoluble residue test used in this research was the test adopted by TxDOT (Tex612-J). The TxDOT AIR test method uses 25 grams of fine aggregates along with aconcentrated hydrochloric acid solution [9]. This test is different from the AIR testdescribed in ASTM D 3042 which uses a 500-gram sample of fine aggregate along with adiluted (6N) hydrochloric acid solution [10].All fine aggregates were also tested using the micro-Deval abrasion test. The micro-Devaltest (ASTM D 7428) is an aggregate durability test that was extensively investigated andrefined by the Ontario Ministry of Transportation. The test consists of placing a pre-soakedaggregate sample (washed and graded) in a jar with a fixed volume of water and a fixedquantity of steel ball bearings. The unit is then put into rotation for a specified period oftime or number of cycles. After the sample is run in the device, it is washed over a No. 200sieve and the retained sample is oven dried. The percent loss in mass is computed fromthe oven dried sample. Aggregates with a low percent loss are considered to be moredurable than the aggregates with a higher percent loss [11].To be able to evaluate the polish resistance of concrete in the laboratory, a method ofsimulating abrasion due to traffic was needed. A Three-Wheel Polishing Device (TWPD)developed by the National Center for Asphalt Technology (NCAT) to test asphalt concretewas purchased and modified (Figure 2). The TWPD was developed to be used with a CTMand DFT. It polishes a circular path on a laboratory specimen that has the same diameteras the path evaluated by the CTM and DFT. The NCAT polisher is composed of threewheels that rotate on a laboratory specimen for a specified number of cycles. Circular ironplates can be placed on the turntable to change the weight on the TWPD, and, hence, thestress on the concrete. The TWPD also has a water spray system that sprays water on thesurface being polished. NCAT added the water spray system to wash away the abradedparticles, simulate wet weather conditions, and to extend the life of the wheels. Themodifications made to the NCAT device included changing the wheels from pneumatic topolyurethane (durometer of 85) and adding a vibration dampener. Each set ofpolyurethane wheels was replaced after 500,000 polishing cycles. The stress caused bythe wheels on the concrete surface was estimated to be around 50psi (based on the totalload and the contact area).Figure 2 – Modified NCAT Three-Wheel Polishing DeviceSURF1-Rached5

Two slabs 20 in. wide and 3 ½ in. deep were tested for each mixture. The change intexture and friction was monitored over 160,000 polishing cycles using the CTM and DFT.Measurements were taken initially and after 5,000, 40,000, 100,000, and 160,000polishing cycles. To evaluate the same polished area, each slab was marked so thatreadings could be taken at the same location. All slabs were textured using a broom finish,and the surface was cured for at least 28 days before the slabs were tested. Two texturereadings were measured using the CTM for each slab at each polishing interval, theprocedures described in ASTM E 2157. When measuring friction using the DFT, ASTM E1911 reports that standard deviation on the same test surface for DFT60 is 0.038, and forthis reason friction measurements using the DFT at 40,000, 100,000, and 160,000 cycleswere repeated several times on the same slab at the same location until the differencebetween the last two readings was less or equal than 0.01. The last measurementobtained (usually the lowest) was reported. Figure 3 shows a picture of the wear patternproduced by the TWPD on a concrete slab.Figure 3 – Typical Wear Pattern Produced by the TWPD on a Concrete SlabThe CTM and DFT were also used to evaluate field sections. Measurements in the fieldwere taken in the wheel path and between wheel paths. The measurements taken in thewheel path represent the abraded concrete (current condition), while the measurementstaken between the wheel paths are good estimates of the original condition of thepavement before it was subject to traffic.4. TEST RESULTS AND DISCUSSION4.1. Field TestingFive field sections in two different locations in the Ft. Worth district were evaluated. Thosesections were chosen because they were the only known sections in Texas that weremade with materials that did not meet the TxDOT AIR limit of 60%. The first location hadtwo sections that were constructed with 100% limestone MFA, while the second locationcontained three sections made from three different blends of siliceous sand and limestoneMFA. The difference between the two sections made with 100% MFA (AIR 0%) was ingradation, not source; section 1 had 5% aggregates passing the No. 200 sieve(microfines) while section 2 had 10%. Those two sections were constructed in 2008 aspart of an implementation project that involved using manufactured fine aggregatescontaining high microfines. The other three sections were constructed in 1995 usingblends of sands that do not meet the 60% AIR limit (AIR of 29%, 35%, and 40%). All fivesections experience truck traffic; the exact traffic count was not obtained. The resultsshown in Figure 3 are average equivalent skid values that were computed using theSURF1-Rached6

equations 1, 2, and 3. These values are average of three measurements taken on thewheel path of each of the section evaluated. The value obtained for the 100%manufactured limestone between wheel paths were double the values shown for the wheelpath in Figure 3. The blended sand sections had values much higher than those of thesections made with 100% MFA. SN(40)ribbed increased as the siliceous content (or AIR) ofthe blended sand in the pavement increased.Note that the values presented in Figure 3 represent calculated skid numbers and notactual skid numbers obtained using a skid trailer. State agencies do not normally use suchmethods to evaluate skid resistance on pavements. TxDOT for example, tests pavementsusing a skid trailer at 50 mph (80 km/h) using smooth tires and not ribbed tires.Figure 4 – SN40 for Test Sections made with 100% MFA and Blended SandsThe values presented in Figure 4 only compared the PCC pavement sections based onthe type of sand that was used, and based on the age of the sections. To be able to do abetter comparative analysis of those sections it is important to also compare how muchtraffic each of those sections is exposed to. Although traffic data were not used for thecomparison, the data presented in Figure 4 are sufficient to show that there is significantperformance difference between PCC pavements made with blended sands and PCCpavements made with 100% manufactured limestone sands.4.2. Laboratory TestingAfter 160,000 TWPD polishing cycles, the change in macro-texture (MPD measured by theCTM) was minimal. The friction value obtained from the DFT after 160,000 polishingcycles at 60 km/h (DFT60) was used to compare the performance of the fine aggregatestested (the DFT evaluates the micro-texture). Note that DFT60 and DFT20 are commonlyused to compare results obtained using the DFT. Results for all the sands and blends ofSURF1-Rached7

sands tested are shown in Figures 5, 6 and 7. The average difference in DFT60 valuesbetween two slabs made from the same material at 160,000 cycles was 0.006.Results presented in Figure 5 show that all the limestone sands tested (except for one)had DFT60 values after 160,000 polishing cycles that were lower than any of the siliceoussands tested. The dolomite sand had a DFT60 value at 160,000 cycles that wascomparable to the values obtained with siliceous sands. Unlike the one limestone thatperformed well, the dolomite was expected to perform better than the other carbonateaggregates. Laboratory results obtained by Balmer and Colley also showed that dolomiticsands had higher wear indices compared to limestone sands [3]. Balmer and Colley’sconclusions did not reflect the differences between carbonate sands because they basedtheir conclusions on AIR. AIR cannot differentiate between carbonate sands since allcarbonates sands regardless of their hardness dissolve in acid. No field sectionscontaining the dolomite sand tested could be located (since that sand does not meet AIRrequirements). It is not clear whether or not dolomite sands would have good performancein the field as they had in the laboratory.FIGURE 5. DFT60 Results at 160,000 Cycles for the Different Sands TestedCombinations of limestone and siliceous sands were blended and tested in concrete. Theresults presented in Figures 5 and 6 show that even the concrete made with aggregateblends that had an AIR of 20% performed considerably better than the concrete made with100% manufactured limestone sand. Adding a small quantity of siliceous sand had a largeeffect on friction performance. The results also indicate that increasing siliceous sandcontent resulted in higher friction values after 160,000 polishing cycles. Those results arealso similar to the results obtained by Balmer and Colley [3].SURF1-Rached8

Figure 6 – DFT60 Results at 160,000 Cycles for Blended Sands (Siliceous 1 andLimestone 1)Figure 7 – DFT60 Results at 160,000 Cycles for Blended Sands (Siliceous 2 andLimestone 2)In Figure 8, the results shown in Figures 5, 6 and 7 were compared to the AIR values.Except for the carbonate sands, there does seem to be a linear relationship between AIRand the friction values obtained for all sands and blends of sands tested. As AIRdecreased, skid performance decreased for the siliceous sands and the blended sands.SURF1-Rached9

FIGURE 8 – DFT60 at 160,000 Cycles vs. AIR (%)In Figure 9, the results shown in Figures 5, 6 and 7 were compared to micro-Deval percentloss of the sands (ASTM D 7428). Except for one of the limestone sands, the micro-Devalabrasion test seems to better predict the performance of the laboratory concrete specimentested. In general, very good friction performance can be expected with aggregates (orblends) that have a micro-Deval percent loss less than 12%.FIGURE 9 – DFT60 at 160,000 Cycles vs. Micro-Deval (%)Figure 10 shows that there is good correlation between AIR values and micro-Devalvalues for all aggregates except one. The only aggregate that did not perform well in AIRbut had good micro-Deval performance was the dolomite sand. Dolomites are known to beharder carbonate aggregates, and the reason they fail AIR is because they are carbonatesand not because they are soft. The 12% micro-Deval limit also seems to correlate well withthe 60% AIR limit; the only major difference between the two limits is that if the micro-SURF1-Rached10

Deval limit was adopted, it would allow more of the harder carbonate sands to be used (orblends containing higher amounts of manufactured carbonate sands).FIGURE 10 – AIR (%) vs. Micro-Deval (%)If blends of the siliceous and limestone aggregates tested during this research projectwere to be blended to meet a micro-Deval loss of less or equal to 12% (MD 12%), thenthe minimum AIR that can be obtained from such blends would be greater than 40%(Figure 11). A field section containing a blend of aggregates with an AIR of 40% wasevaluated as part of this project. That section seemed to have maintained goodperformance after 15 years of service (Figure 4).FIGURE 11 – AIR Values for Blends of Aggregates Meeting the 12% Micro-Deval LimitSURF1-Rached11

5. CONCLUSIONSGood quality concrete can be produced using MFA if the aggregates are properlyevaluated and the right proportions are used. Using 100% limestone sand is notrecommended because it might cause early loss of skid resistance. To obtain good skidperformance using limestone sands, these sands have to be blended with siliceous sands.For a given sand combination, the higher the siliceous sand content, the better the longterm skid performance.The acid insoluble residue test used by state agencies is a surrogate test that measuresthe carbonate content of fine aggregates. Although higher carbonate content mightindicate the presence of softer sands, the AIR does not directly measure the hardness ofthe aggregate. The micro-Deval test, on the other hand, is a mechanical test that directlyevaluates the hardness by measuring the abrasion resistance of the aggregate. Theresults obtained in this research indicate that the micro-Deval abrasion test for fineaggregates is more reliable for predicting the performance of concrete tested at thelaboratory compared to the AIR test.Although only one dolomite sand was tested, dolomitic aggregates are expected toperform better than limestone aggregates. Since dolomite aggregates have not beenextensively evaluated in the field, it is hard to recommend using 100% dolomite fineaggregate in concrete paving. However, it would be safe to recommend using more of aharder carbonate aggregate in a blend. Using at least 25% siliceous sand asrecommended by Balmer and Colley should not cause early loss in skid resistance, but itwill probably not perform as well as a blend containing higher percentage of siliceoussand. Using an AIR limit of 60% or a micro-Deval percent loss of 12% should result ingood skid performance in PCC pavements.REFERENCES1. Folliard, K.J., and K.D. Smith (2003). Aggregate Tests for Portland Cement Concrete Pavements Reviewand Recommendations. National Cooperative Highway Research Program, Research Report No. 281,Transportation Research Board of the National Academies, Washington, D.C.2. Hall, J.W., K.L. Smith, and P. Littleton (2009). Texturing of Concrete Pavements. National CooperativeHighway Research Program, Research Report No. 634, Transportation Research Board of the NationalAcademies, Washington, D.C.3. Balmer, G.G., and B. E. Colley (1966). Laboratory Studies of The Skid Resistance of Concrete. PortlandCement Association. Research and Development Laboratories, Development Department, BulletinD109.4. Renninger, F.A., and F.P., Jr. Nichols (1977). Aggregates and Pavement Skid Resistance. GeologicalSociety of America Engineering Geology Case Histories, No. 11, pp. 25-29.5. Jackson, N.M. Harmonization of Texture and Skid Resistance Measurements (2008). FloridaDepartment of Transportation Research Report, Fl.DOT/SMO/08-BDH-23, University of North Florida,College of Computing, Engineering and Construction 1 UNF Drive Jacksonville, FL 32224.6. Hall, J.W., L.T. Glover, K.L. Smith, L.D. Evans, J.C. Wambold, T.J. Yager, and Z. Rado (2006). Guide forPavement Friction. National Cooperative Highway Research Program, Research Report No. 108,Transportation Research Board of the National Academies, Washington, D.C.7. PIARC Technical Committee on Surface Characteristics (C1) – (1995). International PIARC Experimentto Compare and Harmonize Texture and Skid Resistance Measurements. Paris, France.8. ASTM E 1960 (2007). Standard Practice for Calculating International Friction Index of a PavementSurface. American Society for Testing and Materials, Philadelphia, PA.9. Tex-612-J (2000). Test Procedure for Acid Insoluble Residue for Fine Aggregate Texas Department ofTransportation, Austin, TX.10. ASTM D 3042 (2009). Standard Test Method for Insoluble Residue in Carbonate Aggregates. AmericanSociety for Testing and Materials, Philadelphia, PA.SURF1-Rached12

11. Rogers, C. A., M. L. Bailey, and B. Price (1991). Micro-Deval Test for Evaluating the Quality of FineAggregate for Concrete and Asphalt. Transportation Research Record: Journal of the TransportationResearch Record, No. 1301, Transportation Research Board of the National Academies, Washington,D.C., pp. 68-76.SURF1-Rached13

The Dynamic Friction Tester (DFT) is an apparatus that measures the friction-speed relationship on a pavement surface for speeds ranging from 0 to 80 km/h (micro-texture). The DFT measures the torque needed to stop three small spring-loa