Long and short term stability of straight and polymer modified asphalts.

In recent years various studies have shown that polymer modification can be a successful avenue in obtaining asphalt binder systems with improved property sets. Improvements were noted in the binder properties, the properties of the hot mix asphalt (HMA) as well as in actual pavements in the field. In the case of permanent deformation Valkering, et al (ref. 1) found that in wheel tracking experiments the rutting rate at 40 [degrees] and 50 [degrees] C could be significantly reduced via polymer modification. These results have been substantiated by Bouldin and Collins (ref. 2) with the TRRL wheel tracker for temperatures up to 60 [degrees] C. Both repetitive and static creep experiments on HMA cores appear to correlate with these findings (refs. 1, 3-5). Reports by Reese, et al (ref. 6) and Fleckenstein, et al (ref. 7) for example show that on heavily trafficked pavements dramatic improvements have been observed. In the case of thermal and fatigue cracking both Collins, et al (ref. 8) and Goodrich (ref. 9) have observed enhanced performance.

Such results have lead to wide-spread commercial use of polymer modified asphalts. In some states, for example Nevada, polymer modification of asphalt is currently being specified for wearing courses on all major thruways. We will elucidate how the long and short term stability of both conventional and polymer modified asphalt influences their performance. The issues we will address are:

* asphalt cement aging;

* polymer stability;

* polymer-asphalt phase stability and microstructure;

* and how these influence the binder properties, as well as the actual field performance.

The way these materials are handled in the field, i.e., from initial polymer-asphalt blending to lay down and compaction can have a significant impact on the previously mentioned issues. The first section will therefore primarily address short term aging encountered in these operations. In the second part, we will discuss how long term field aging can affect the pavement performance.

Experimental

Sample preparation, microscopy, rheological characterization and traditional binder tests are described in detail in reference 10.

Gel permeation chromatography

A gel permeation chromatography (GPC) technique was used similar to that described by Portfolio and Fensel (ref. 11) to determine mean molecular (Mw) weight and molecular weight distribution. For the evaluation of field samples, the binders were extracted directly from the HMA with tetra-hydrofuran (THF). The effective amount of polymer is defined as: [Mathematical Expression Omitted]

The unaged samples have therefore per definition an effective polymer concentration of 100%.

Materials

Tables 1 and 2 show the asphalts and polymers used in this study.

     Table 1 - asphalts used in study
Code     Asphalt source    Asphalt grade      Location
NJ       Venezuela            AC-20           NJ Rt. 35
TX       W. TX interm.        AC-10          TX Hwy. 287
WY       Wyoming              AC-10          WY 1-80
NV06     Calif. Valley       AR 1000         NV 1-15
NV0L     Calif. Valley       AR 4000         NV Rt. 95
CA-N1    Calif. Valley       AR 1000         CA I-40
CA-N2    Calif. Valley       AR 4000         CA I-40
CA-C1    Calif. Valley       AR 1000         CA Rt. 395
CA-C2    Calif. Valley       AR 2000         CA Rt. 395
CA-S1    Calif. Valley       AR 1000        Sacramento, CA
CA-S2    Calif. Valley       AR 8000        Sacramento, CA
CV       Calif. Valley       AR 4000
ET-1     East Texas           AC-5
ET-2     East Texas           AC-20
VEN-1    Venezuela            AC-7
VEN-2    Venezuela            AC-70
     Table 2 - polymers used in study
Code      Polymer      Oil         Trade name
           type       content        and grade
A           SBS                      Shell Kraton
                                   Rubber D1101
B           SBS          29%          D4141
C           SBS          50%          D4460
D       [(SB).sub.x]                  D1184
E           SEBS                      G1657
F           SEBS                      RP6904
G           SBR        70% solids     Goodyear
                       (no oil)       Ultrapave 70
H         PE                          Dow PE 2045

Low temperature thermal cracking test

We have devised a method to determine the critical cracking temperature (Tcr) of binder systems which is very similar to Hills' method (ref. 12). It is basically a visual method where Tcr is defined as the temperature where the first crack in the asphalt is observed.

ASTM concrete test blocks (7.6 cm x 12.7 cm x 2.4 cm thick) were used as the substrate. These were dried prior to use for 72 hours in a vacuum oven at 100 [degrees] C. The blocks were then cooled down at room temperature. The asphalt cement (AC) was heated to 160 [degrees] C together with the wet film applicator. The asphalt was then removed from the oven, and allowed to cool down to 125 [degrees] C [+ or -] 5 [degrees] and a film of 4 mm thickness (10 to 20g asphalt) spread on the concrete block using the wet film applicator. After cooling to room temperature the sample was placed in an environmental chamber (which had a temperature of 5 [degrees] C). The samples were then equilibrated for an additional 12 hours at 5 [degrees] C before ramping the temperature down at a temperature gradient of 5 [degrees] C/h. Repeatability tests have shown that for both modified and unmodified asphalts the standard variation was better than 1 [degree] C. The cooling rate was found to have no influence, i.e., reduction of the temperature ramp to 2.5 [degrees] C/h yielded the same results.

Repetitive creep experiments

To determine the ability of a certain mix to resist permanent deformation, repetitive creep experiments were carried out. Contrary to some earlier work done by other groups (ref. 1), we did not apply a square wave but used a Haversine wave. The reasons for this are as follows:

* It is easier to control this form of force application. The initial slope dF/dt is very small, and therefore significant force overshoots do no occur.

* This pressure profile tends to mimic much closer the pressures that a pavement will experience in the field ref. 13). The following pressure profile was chosen:

P(t) = [P.sub.0] + [P.sub.1] {1 + sin[n(2t/t.sub.0] - 1/2)]}

with: [P.sub.0] = 2psi   13.8kPa (2psi)
      [P.sub.0] + 2P1 =       551.6kPa (80psi)

[t.sub.0] = 0.365s

The static pressure, [P.sub.0], is required to ensure that there is at all times, contact between the specimen and the platens (the samples accumulate deformation during the test). The resulting ampltitude of the Haversine wave is 537.8kPa (78psi) which is close to the average pressures commonly used in truck tires. No confining pressures were applied.

The actual measurements were carried out on an MTS using a 2,250kg force cell. Between every loading the specimen was given a rest period of 1s to relax. Maximum strain, accumulated strain and recoil were then measured for a period of 5,000 loadings or until the sample had accumulated 6% permanent deformation. All measurements described here were carried out at 40 [degrees] C.

The specimens were prepared from the loose mix by compacting them with the Texas Gyratory Compactor to the required density. (Target air void content was chosen to match in-place field density.) The samples were prepared to a height of 20.32 cm and a diameter of 10.16 cm. In order to obtain smooth parallel surfaces the samples were then cut and polished down to a height of 15.24 cm

     Table 3 - GPC results on polymer modified binders
Asphalt   Polymer      Hot mix     Conditioning    [C.sub.eff%]
                      plant type
NJ        6%w B(1)     Batch       Original(2)             96
                                   RTFO                   82.3
                                   loose mix(3)           84.5
TX        3%w A        Drum        Original               89.3
                                   RTFO                   51.0
                                   loose mix               0.0
          6%w A        Drum        Original               70.6
                                   RTFO                   40.1
                                   loose mix              14.8
WY        6%w C(1)     Drum        Loose mix               100
                                   field aged sample
                                   (after   4 years       87.5
NV01      6%w C(1)     Drum        Original               100
                                   RTFO                   62.9
                                   loose mix              89.9
                                   field aged sample      80.8
                                   (after   3.5 years)
CA-N1       D          Drum        Original               74.9
                                   CATOD                   0.0
CA-S1     4%w F        Drum        Original               95.2
                                   RTFO                   93.4
                                   CATOD                  93.2
                                   loose mix              96.4
(1) Actual heat polymer content
(2) Polymer/asphalt blend prior to use at hot mix plant
(3) loose mix samples were taken prior to compaction at
job site

Short term aging

In this section we would like to focus our interest on how the following processes may affect the performance of both straight and polymer modified asphalts. These process steps are:

* blending of the polymer with asphalt;

* storage of straight and polymer modified asphalts; and

* processing of the binder in a hot mix plant.

Asphalt age hardening

During the processing and storage of straight unmodified asphalt, one usually observes a so-called age hardening of the asphalt (ref. 14). Oxidation leads to more structured asphalts. In general this will result in both a higher asphaltene content and a higher average molecular weights (refs. 15-17). An example of age hardening is shown in figure 1. The asphalt CA-S1 was stored over a period of five days at a temperature of 180 [degrees] C (low shear agitation). As reflected in the penetration at 25 [degrees] C and the penetration at 4 [degrees] C the material becomes significantly harder with time. This effect is even more pronounced in crudes which are more susceptible to aging such as for example heavy Venezuelan crudes (ref. 9).

The most significant asphalt age hardening is normally observed when the asphalt is mixed with the aggregate in the hot mix plant. While exact temperatures are not available, the surface temperature of the aggregate can reach temperatures in excess of 260 [degrees] C. Due to the fact that the AC forms a thin layer on the aggregate, oxidation can be very significant. Thus, most asphalts exhibit an aging index (A.I.) greater than two. This aging index is defined as:

A.I. = [Eta](RTFO)/[Etha] (original)

RTFO stands for rolling thin film oven and is a laboratory aging procedure (ASTM D2872) which is supposed to simulate the aging that is commonly observed in a hot mix plant (ref. 18). Some Western states have adopted this technique and grade all their asphalts according to the viscosity of the aged residue (AR grading system, ASTM D3381).

Polymer stability

Under normal conditions, no significant changes in Mw are observed during the blending process for SBS polymers. This has been established by comparing the molecular weight distribution of the neat polymer with the polymer in modified asphalt. The drop in [C.sub.eff] due to mechanical and thermal conditions in the blending process is generally less than 4 to 5% (c.f. table 3). This is valid for both low and high shear mixing. However, sufficient temperature control is necessary to avoid excessive viscous dissipation during high shear mixing.

At the polymer levels used in standard paving applications, virtually only chain scission is observed. Gelation does not occur (no ultra high molecular weight species even after RTFO). A potential exception is when a concentrate is formulated which may contain up to 15%w polymer. In these cases recombination and crosslinking can result in ultra high molecular weight species and the latter may precipitate during solvent extraction.

A more likely source of polymer degradation is hot storage. It is not uncommon in the industry to hold asphalts over periods of weeks, and less frequently over months. If this is done with styrene butadiene (unsaturated) polymers, one should take care to ensure that the storage temperature is well below 135 [degrees] C to maximize polymer stability. Two lab samples were blended both containing 4%w SBS. The first one was held at a temperature of 180 [degrees] C (356 [degrees] F) and the second sample was held at a more moderate temperature of 120 [degrees] C (248 [degrees] F). In both cases the samples were gently agitated and were exposed to air. As shown in figure 2 at 120 [degrees] C, the polymer stability is significantly enhanced relative to storage stability at 180 [degrees] C. Stability can be further improved by blanketing with nitrogen.

This stability has a potentially significant impact on the resulting rheological properties of the asphalt blend, and thus its relative performance in the field. In general one would want a material that exhibits high resistance to deformation at elevated temperatures and that is not prone to exhibit cracking at low temperatures. A measure of the ability of a material to resist deformation is the complex modulus, G*. The elasticity of a material is measured with the storage modulus, G', which indicates the portion of the energy that is restored elastically. The loss modulus, G", is an indicator of the amount of energy that dissipates in the form of viscous flow (ref. 19). Straight unmodified asphalts are at elevated temperatures virtually only viscous,

G" >> G'

G" - G* while polymer modified asphalt may be even more elastic than viscous at high temperatures (refs. 2 and 10). The ideal binder should have a large value of G* and G' at 60 [degrees] C.

On the other hand, to prevent premature fatigue cracking G* at 25 [degrees] C should be as low as possible (refs. 20-22). In figure 3 the values of G* and G' are plotted as a function of storage time for the two samples of the asphalt CA-S1 containing 4%w SBS (polymer D). It is interesting to note that the sample stored at 180 [degrees] C exhibits a rather significant drop of G' over the storage period while the sample stored at 120 [degrees] C shows a much smaller decrease in elasticity. Fragmentation of the SBS copolymer to lower molecular weight species thus results in a reduction of the material's ability to resist permanent deformation. By the same token the complex modulus of the material at 25 [degrees] C increases when stored at 180'C. Hence, polymer modified asphalt stored at an excessively high temperature of 180 [degrees] C will be more susceptible to fatigue cracking and thermal cracking due to asphalt age hardening (high G*). Best overall property retention is achieved by storing the samples at temperatures not exceeding 120 [degrees] C and by minimizing the duration of storage.

As previously mentioned, the RTFO is a standard testing technique to approximate what will happen to the material in the hot mix plant. In table 3 a series of SBS and SEBS modified asphalts were compared to see if RTFO is a good indicator for polymer degradation. In general we find that the RTFO is more severe to the polymer than the "real-world" hot mix plant. In most cases RTFO significantly over-predicted polymer stability. This appears to be independent of the plant type, continuous or batch. Additionally, saturated polymers (polymer F and H) show virtually no change following either RTFO, long term storage, or hot mix plant processing. Thus we conclude that RTFO aging is probably only an indicator if particular polymer asphalt systems are potentially unstable.

In the case of the SEBS modified asphalt, RTFO aging was more severe than that observed in the loose mix (c.f. table 3, CA-S1 4%w polymer F). The degradation was, however, for SEBS insignificant and led actually to enhanced high temperature properties of the binder (c.f. CA-S1 4%w polymer F). It is interesting to note that only one of the six SBS samples showed a decrease in elasticity at 60 [degrees] C after RTFO. This reduction of G' is a direct result of polymer fragmentation. Asphalt age hardening, on the other hand, generally gives an increase in G* (and a decrease of the penetration). In all the loose mix samples in this study no reduction of G' or G* was observed, again indicating that RTFO is more severe than the actual processing.

Morphology of polymer blends

A number of papers have dealt with the characterization of the morphology of polymer modified asphalts. Both fluorescence (refs. 23-25) and electron microscopy (ref. 10) have been used. As we had pointed out in an earlier paper, fluorescence microscopy is not necessarily in all cases the analytical tool of choice. Moreover, it only yields a relatively coarse picture of the fluorescence active areas. As some of the components in the asphalt are very fluorescent, namely the polar aromatics and naphthenics, one obtains a more precise notion of the underlying morphology using scanning transmission electron microscopy (STEM) or environmental scanning electron microscopy (ESEM). The latter is especially interesting as bulk samples with little preparation can provide images with wider fields of view than STEM.

In order to understand what influence processing may have on the morphology, we looked at SBS, SEBS, SBR and PE in the asphalts CA-S1 and ET-1. Photomicrographs were made of pre- and post-RTFO samples. In the case of SBS (polymer A) in the relatively compatible asphalt ET-1 we observed no changes in the morphology. Particle size and particle size distribution remained unaffected. SBR (polymer G) on the other hand, showed a coarsening of particles which may explain the reduction of G' in the RTFO aged SBR samples. However, polymer degradation may account to some degree for this reduction of G'. Unfortunately, this GPC technique could not be used to determine the degree of degradation in polymer G.

Results for SEBS (polymer F) in asphalt CA-S1 show a finer dispersion after RTFO. In this case RTFO actually provides additional mixing and is not detrimental to the rheological properties of the blend. Because of the very coarse structure of PE (polymer H) modified asphalts, electron microscopy was not appropriate. The fluorescence photomicrographs show that gross phase separation occurs during the RTFO aging process. This is reflected in the remarkable reduction of both G* and G' after RTFO ( 90% decrease) which is not due to polymer degradation.

It is difficult to ascertain the actual microstructure in the finished pavement Certainly each step in processing will impact the resulting structure. For example, the heat history and mixing conditions will be of importance. Static tests such as the can test are not a realistic simulation and can be misleading. Perhaps ESEM may prove to be a suitable non-obtrusive technique to study the morphology of polymer/asphalt blends in HMA

Case histories - short-term aging

New Jersey State Route 35

In November 1987 New Jersey placed test sections on Rt. 35 in Hazlet Township at heavily trafficked intersections. Rutting and shoving were the major concerns.

As shown in table 1 a heavy Venezuelan crude (AC 20) was chosen as base asphalt. The traditional binder properties of the polymer modified AC are shown in table 4 Polymer B has 29%w extender oil. When blended with a straight asphalt this oil will always lead to a softer base asphalt. An excellent way to track how the oil influences the base stock is the 4 [degrees] C pen. This stems from the fact that the low temperature stiffness of the binder is virtually unaffected by polymer addition. Table 5 shows an example for asphalt ET-1 modified with SBS (polymer A). Over a concentration range from 0w to 6%w polymer the 4 [degrees] C pen remains constant. On the other hand the polymer already has an effect on the modulus of the material at 25 [degrees] C and the pen at this temperature decreases with increasing polymer concentration. Comparison of N.J. with 6%w (neat) polymer B with VEN-2 shows that the 4 [degrees] C pen has significantly increased due to the oil addition (from 18 to 24dmm on the RTFO aged sample). [TABULAR DATA 4 OMITTED]

     Table 5 - influence of SBS (polymer A) on penetration in asphalt ET-1 (AC-5
)
%w Polymer   25 [degrees] C pen   4 [degrees] C pen
   0               150                  35
   1.4             133                  34
   2               128                  34
   3               118                  35
   4               116                  36
   5               103                  35
   6                97                  35

This oiling has a very significant influence on the properties of the binder system. It will make the binder less stiff at low temperature and thus improve the ability of the system to resist thermal and fatigue cracking. On the other hand it will lead to a softer, less asphaltene rich binder which will be more compatible with styrene-rubber copolymers such as SBS, SBR and SEBS. This will consequently lead to a more pronounced network formation. The latter will improve the high temperature properties of the binder, i.e., its resistance to permanent deformation. The fact that soft, compatible asphalts will lead to more elastic binders with improved rutting resistance is discussed in more depth by Bouldin, et al (ref. 10).

The GPC results shown in table 3 indicate excellent polymer stability in both the RTFO aged sample and the loose mix sample. These results imply that:

* the base asphalt did not promote polymer degradation

* RTFO appeared to simulate hot mix plant processing very well in this case.

For this job the HMA was made in a batch plant. In general we find that batch operations affect polymer/-asphalt stability less than continuous drum plants. The reason for this is that the latter show greater temperature variability. An example will be shown in the following case history. However, as shown in table 3 excellent results can be obtained in drum plants if process controls are in place to ensure that the HMA temperature does not exceed

160 [degrees] C (c.f. table 3 WY).

The rheological characterization of the recovered binder (ASTM D1856) suggests that this binder should provide good resistance to permanent deformation. Both the complex modulus and the storage modulus are very high at 60 [degrees] C The results support the notion that the binder in the loose mix is superior to the RTFO aged residue. The reason for this behavior may be at least in part attributed to additional mixing afforded by the pug mill. On the other hand the extraction and subsequent recovery may lead to morphological changes. As previously mentioned the morphology has in many cases a prominent influence on the rheological properties of the modified asphalt. Additional work will be required to verify the general applicability of this extraction method for polymer modified asphalts.

In order to determine the relative contribution of the binder to mixture performance repetitive creep experiments were carried out and 5,000 loading cycles were required to achieve 6% deformation. The AC-20 unmodified control, on the other hand, failed after only 1,200 cycles.

Similar results have been reported in terms of field performance. The AC-20 has already exceeded N.J.'s limits for rutting (1.27cm rut depth) after three years. Rut depths in the polymer modified test sections were found to be between 0.64 to 0.95cm (1990).

Texas Highway 287

In the spring of 1989 Texas constructed HMA test sections using various polymer modified asphalts on Hwy. 287 near Memphis, Texas. Two test sections were put down containing 3%w and 6%w SBS (polymer A). The base asphalt was as shown in table 1, and AC-10 (TX). However, in the case of the binder containing 6%w polymer, an extender oil was added. Hence, the resulting base asphalt was, according to table 4 (c.f. TX 6%w polymer A 4 [degrees] C pen), relatively soft and resembled more an AC-5 (c.f. table 5).

Infrared pyrometer measurements on the 3%w polymer A section taken on the mat directly behind the paver gave excessively high temperatures, 180 [degrees] C. Normal mat temperatures should be in the neighborhood of 150 [degrees] C or lower. The results of the excessively high temperatures can be seen in the material properties. The values of both the complex modulus and the storage modulus of the original binder can be compared with those of the binder recovered from the loose mix and the more moderate increases observed due to RTFO. The increase in G* at 25 [degrees] C is almost an order of magnitude, from 67,800 to 615,000 Pa. This value is actually higher than values typically found for an AC-30 grade asphalt (ref 8), i.e., the asphalt had dramatically age hardened. As one may expect under these severe conditions the polymer/asphalt blend was largely degraded. No triblock material could be detected by GPC for the blend containing 3%w SBS. For the blend containing 6%w SBS serious stability problems were observed and Ceff was reduced to 14.8%. In this particular case RTFO predicted a significantly higher degree of polymer stability (c.f. table 3). However, note that the absolute amount of degradation is nonetheless unusually high in this asphalt. This implies that certain asphalts lead to less stable polymer/asphalt binder systems. Hence, RTFO coupled with GPC and rheology is a good tool in predicting polymer stability.

The results for repetitive creep are actually slightly higher for the binder system with less polymer. The reason for this is that polymer content has virtually no influence on the performance due to the high degree of degradation in both instances. Moreover, the base asphalt stiffness is in these cases the over-riding factor.

On the other hand, potential premature rutting could have occurred if the polymer degradation would not have been accompanied by significant asphalt age hardening. Therefore it is of particular importance when modifying soft base asphalts with low levels of polymers to ensure polymer integrity.

Nevada I-15

In 1989 Nevada placed a series of polymer modified asphalts on 1-15 near Mesquite. They were seeking improvements in resistance to permanent deformation and long-term pavement durability which are common problems throughout the deserts of the Southwestern U.S. In these regions the average daily high temperatures of the hottest month exceed 400C and pavement temperatures up to 80 [degrees] C have been measured by NDOT (ref. 26).

The asphalt binder system consisted of a California Valley AR-1000 (soft base; NV06) modified with polymer D at 2.5%w and 3.8%w SBS. The binders exhibit very high elasticity (G') and resistance to deformation (G*) at high temperatures. In addition, low temperature properties are outstanding.

This is reflected in the relatively low values for G* at 25 [degrees] C (soft material is less susceptible to fatigue cracking) and the very low critical cracking temperatures. This best balance of properties results from using a soft base asphalt and sufficient polymer to establish a polymeric network. The effectivity of this network formation is demonstrated by the exceptionally high numbers of loadings until failure in repetitive creep of the HMA. Increasing the polymer concentration from 2.5%w to 3.8%w more than doubles the resistance to permanent deformation of the HMA. A more detailed description of how the rheology of these binders influences their performance is given in references 2 and 27.

Sacramento Mack Road/Stockton Blvd.

The City of Sacramento was concerned about the severe shoving and rutting at the intersection of Mack Rd. and Stockton Blvd. In order to evaluate how an SEBS modified asphalt would perform in comparison to a very hard AR-8000 (CA-S2) two pavements were laid down in October of 1990.

Virtually no degradation was observed in the original, the RTFO aged residue and in the extracted binder (cf. table 3). In general saturated polymers are relatively immune from degradation problems. Therefore, plant operating windows can be much wider. Yet, good handling practices are still imperative to avoid the conesequences of serious asphalt age hardening.

With regard to the mechanical properties of the binder systems it is worth noting that the polymer modified asphalt exhibits superior performance at both high and low temperatures, i.e., higher values of both G* and G' at 60 [degrees] C and lower values at 25 [degrees] C. The critical cracking temperature of the SEBS modified material is significantly lower than that of the AR 8000. The pavement is still too new to draw any conclusions.

Long term aging

Following lay-down and compaction there has to be concern with the long-term durability of the pavement and how this impacts the performance of the HMA. This aging process is exacerbated by the severe desert-like climates which prevail throughout large areas of the Southwestem United States.

In the case of straight unmodified asphalts the experience in these areas has been that ACs embrittle within two to four years. Subsequently the pavements fail due to cracking. In permeable mixes, for example open graded friction courses or gap graded mixes, aging may be of concern even in moderate climatic regions. A possible avenue in obtaining long-term aging resistant binder systems is polymer modification. Polymer modification allows the use of soft base asphalts and, therefore the bitumen itself will not govern the viscoelastic behavior of the blend. Moreover, the polymeric network is the determining factor. Therefore, polymer stability is of critical concern.

Various attempts have been made to devise laboratory simulations for long-term aging. A detailed evaluation of simulated age hardening of asphalt mixtures has been published by Chollar, et al (ref. 16) and Chari, et al (ref. 28). Chari reported on the effects of UV and elevated temperatures on the aging rate of a number of straight asphalts. More recently the Strategic Highway Research Program (SHRP) has been evaluating the so-called pressure air vessel (PAV) method (refs. 29 and 30). However, to our knowledge this work has not been extended to include polymer modified systems. Also, further field validation is required.

A simple and straightforward technique to simulate long-term field aging (as it occurs in desert regions) has been developed by Kemp and Predoehl at Caltrans (ref. 31). This test utilizes the standard rolling thin film oven to age the asphalt at a temperature of 111 [degrees] C for seven days. The procedure of this so called California tilt oven durability test (CATOD) is given in California Test Method 874. Both Kemp and Predoehl (ref. 31) as well as Reese and Predoehl (ref. 6) have been successful in correlating field data with CATOD lab results. There is sufficient data currently available to determine the general applicability of this test to other cimatic regions. At present polymer modified asphalts are being evaluated using this technique.

Polymer stability

As in the case of short-term aging, polymer integrity is to be considered in long-term aging. Here also we find GPC to be an excellent tool to track the molecular weight and molecular weight distribution of the block copolymers through time. Two samples were taken from roads which were modified with SBS and are located in Nevada (NV-01) and Wyoming (WY). Both pavements were placed in relatively moderate climates. After three to four years in service we found effective polymer contents of 80.8% and 87.5%, respectively (c.f. table 3) Thus, one can expect a continued contribution of the polymer to the binder performance over an extended period of time.

We also studied the stability of SEBS (polymer F) which due to the saturation of the rubber midblock is expected to give superior long-term performance. Using the previously mentioned CATOD, samples of the binders used in the Sacramento job (CA-S1 with 4%w polymer F and CA-S2) were aged. Virtually no polymer degradation was found (c.f. table 3). In the case of the straight CA-S2 (AR 8000) we found an extreme increase of the moduli. This implies that after some years of service, this pavement is expected to embrittle and exhibit fatigue and thermal cracking. After CATOD aging, the modified material is significantly less brittle and therefore should not experience fatigue and thermal cracking (in this climatic region). Where superior long-term stability is required saturated polymers such as polymer F are the modifiers of choice.

Case histories - long-term aging

In 1987 Caltrans emharked upon a program to find improved binders for desert-like conditions (ref. 6). Some examples will be discussed.

California Interstate 40

For this evaluation Caltrans placed a series of test sections near Needles on 1-40. Needles has a typical low desert climate with an average daily high during the hottest month of 41.7 [degrees] C and an average low during the coldest month of 0 [degrees] C. As a control, an AR 4000 (CA-N2) was placed. Three polymer modified sections were placed in the westbound travel lane, which were separated by control sections. The polymers compared were polymers D, E and G. The base asphalt was AR 1000 in each case. However, as shown in table 4, the low temperature pens indicate that varying amounts of extender oils were used.

To simulate the aging the binders would experience over a period of four years in the field the samples were CATOD aged and subsequently rheologically characterized. The results are shown in table 6. Based on these results, fatigue cracking would be expected to occur with the binders exhibiting the higher moduli at 25 [degrees] C. On a relative basis one would therefore predict the following resistance to reflective and fatigue cracking:

CA-N1/D CA-N1/D > CA-N1/G >> CA-N2

A measure for the binder's ability to mitigate permanent deformation is G* and G' at 60 [degrees] C. The largest potential for rutting is directly after placement before the binder has age hardened. Therefore, it is better to look at the moduli of the unaged material. All materials have a value of G* in the range between 200 and 360 Pa. However, the unmodified asphalt CA-N2 has virtually no elasticity (G' 2) at 60 [degrees] C and, hence, should be much more susceptible to permanent deformation. On a relative basis one would therefore predict the following resistance to permanent deformation.

CA-N1/G CA-N1/D > CA-N1/E >> CA-N2

A summary of the actual field performance is depicted in figure 4. All the control sections are showing reflective and wheel path cracking. As predicted the controls in the heavily tracked driving lane are also showing rutting and bleeding (ref. 32). All polymer modified sections have shown excellent resistance to permanent deformation. Only in the CA-N1/G section has reflective cracking been observed. Thermal cracking has not been detected in any of the sections. The field performance confirms the predictions offered based on the rheological findings on the CATOD aged samples.

California U.S. Route 395

A cold climate test section was placed on U..S. Rt. 395 at Crestview, approximately 50 miles north of Bishop. This test section does not experience the high temperatures that the Needles test sections receive. The average daily high for the hottest month is 32 [degrees] C, according to Caltrans. The winters are colder with an average low for the coldest month of- 15 [degrees] C.

As a control an AR 2000 (CA-C2) was chosen. The polymer modified system was AR 1000(CA-C1) containing polymer E. As shown in table 4, the system contained a significant amount of extender oils which is reflected in the very high 4 [degrees] C pen.

Due to the low traffic volume and the fact that it doesn't experience excessively high pavement temperatures, fatigue and or rutting were not expected. However, thermal cracking should occur in the CA-C2 which, according to table 6, has a critical cracking temperature of - 19.4 [degrees] C after CATOD aging. The polymer modified system has a Tcr of -30.9 [degrees] C after CATOD aging and, therefore, should not exhibit thermal cracking.

The field observations agree very well with the laboratory findings. The control has already shown thermal cracking after the exceptionally cold winter of 1990/91. The polymer modified test section is not showing distress.

Conclusions

Both short and long-term aging can have a significant influence on the performance of HMA. This is true for polymer modified asphalts as well as for conventional AC's. The latter can undergo significant age hardening during storage, mixing and in the field. This long term aging can be dramatic and lead to premature pavement failure. The effects of aging can be exacerbated in the case of permeable pavements or in regions with extreme climatic conditions.

Polymer modification is a viable route in overcoming these potential deficiencies by providing binders with well balanced property sets. This is generally achieved by modifying soft, compatible asphalts. Improvements are noted in both the binder's resistance to permanent deformation and its ability to mitigate thermal and fatigue cracking. However, unsaturated polymers may exhibit thermal and oxidative instabilities. Thus, it is imperative to avoid prolonged storage and excessive temperatures. This is especially important in hot mix plant operations. GPC and dynamic mechanical analysis are found to be excellent tools in evaluating the stability of polymer (SBS) modified asphalts. Likewise RTFO can be useful in indicating potentially unstable polymer/asphalt blends. The results presented here show that different asphalt chemistries can lead to varying degrees of polymer degradation in the case of unsaturated polymers. In general, we found RTFO aging to be more severe than actual field operations. With regard to long-term aging Caltrans and our findings appear to support the notion that CATOD is a reasonable simulation for field aging in hot desert-like climates. In more moderate climates we observed less than 20% reduction of [C.sub.eff] even after three to four years in the field.

Improved thermal and oxidative stability can be achieved by modifying asphalts with saturated polymers. However, good handling practices are still imperative to preclude the consequences of serious asphalt age hardening. GPC, rheological characterization and field data demonstrate the superior performance of these blends even in severe climates

References

[1.] C.P. Valkering, D.J.L. Lancon, E. de Hilster and D.A. Stoker, "Rutting resistance of asphalt mixes containing nonconventional and polymer-modified binders," Association of Asphalt Paving Technologist Proc. AAPT, 59(1990). [2.] M.G. Bouldin and J.H. Collins, "Wheel tracking experiments with polymer modified and unmodified hot mix asphalt," Polymer Modified Asphalt Binders, ASTM STP1108, K.R. Wardlow and S. Shuler, Eds., ASTM, Philadelphia, 1990. [3.] D.N. Little, "Analysis of the influence of low density polyethylene modification (Novophalt) of asphalt concrete on mixture shear strength and creep deformation potential," Polymer Modified Asphalt Binders, ASTM STP1108, K.R. Wardlow and S. Schuler, Eds., ASTM, Philadelphia, 1990. [4.] A.A. Tayebali, J.L. Goodrich, J.B Sousa and C.L. Monismith, "Relationships between modified asphalt binder rheology and binder-aggregate mixtures permanent deformation response," Association of Asphalt Paving Technologist Proc. AAPT, 60 (1991). [5.] R.J. Salter and M.Y.O. AI-Shakarchi, "Effects of ambient temperature and thermal cycling on the creep of bituminous pavement material," TRB Record 88052, Washington, D.C., 1989. [6.] R. Reese and N. Predoehl, "Evaluation of modified asphalt binders," Caltrans Report #65324-633383, Sacramento, 1989. [7.] L.J. Fleckenstein, K. Mahboub and D.L. Allen Performance of polymer modified asphalt mixes in Kentucky," Polymer Modified Asphalt Binders, ASTM STP1108, K.R. Wardlow and S. Shuler, Eds., ASTM, Philadelphia, 1990. [8.] J.H. Collins, M.G. Bouldin, R. Gelles and A. Berker, "Improved performance of paving asphalts by polymer modification," Association of Asphalt Paving Technologist Proc. AAPT, 60 (1991). [9.] J.L. Goodrich, "Asphaltic binder rheology, asphalt concrete rheology and asphalt concrete mix properties," Association of Asphalt Paving Technologist Proc. AAPT, 60 (1991). [10.] M.G. Bouldin, J.H. Collins and A. Berker, "Rheology and microstructure of polymer/asphalt blends," Rubber Chem. Tech. (in print). [11.] D.C. Portfolio and F.A Fensel, "Effect of heat history during manufacture of SBS modified bitumen coating quality," Proc. 3rd Intl. Symp. Roofing Techn., Montreal, 1991. [12.] J.F. Hills, "Predicting the fracture of asphalt mixes by thermal Stress," Inst. Petr. Tech. 74-014 (1974). [13.] J.B. Sousa, J. Craus and C.L. Monismith, "Summary report of permanent deformation in asphalt concrete," SHRP-A/1R-91-104, Washington, D.C., 1991. [14.] J.C. Petersen, "Chemical composition of asphalt as related to asphalt durability: State of the art," TRB Record 999, Washington, D.C., 1984. [15.] A.S. Noureldin and L.E. Wood, "Variations in molecular size distribution of virgin and recycled asphalt binders associated with aging," TRB Record 880271, Washington, D.C., 1989. [16] B.H. Chollar, J.A. Zenawitz, J.G. Boone, K.T. Tran and D.T. Anderson, "Changes occurring in asphalts in drum dryer and batch (pug mill) mixing operations," TRB Record 880219, Washington, D.C., 1989. [17.] J.A.S. Prabanic, M. Emmelin and G.N. King, "Use of a multi-wave length UV-VIS detector with HP-GPC to give a 3-D view of bituminous material," TRB Record 880348, Washington D.C., 1989. [18.] F.N. Hveem, E. Zube and J. Skog, "Proposed new tests and specifications for paving grade asphalts," Association of Asphalt Paving Technologist Proc. AAPT 32 (1963). [19.] J.D. Ferry, "Viscoelastic properties of polymers," Third Edition, John Wiley & Sons, New York, 1980. [20.] S.C.S.R. Tangella, J. Craus, J.A. Deacon and C.L. Monismith, "Summary report in fatigue response on asphalt mixtures," SHRP-A/1R-90-011, Washington, D. C, 1990. [21.] J.B. Sousa, G. Rowe and A.A. Tayebali, "Dissipated energy and fatigue of asphalt-aggregate mixtures," (submitted for publication). [22.] W. van Dyke and W. Visser, "The energy approach to fatigue for pavement design," Association of Asphalt Paving Technologists Proc. AAPT 46 (1977), 1-40. [23.] G. Kraus, "Modification of asphalt by block polymers of butadiene and styrene," Rubber Chem. Tech. 55 (1982), 1389-1402. [24.] C. Lenoble, "Rheology and microstructure of aged polymer modified bitumens," Proc. of Inter. Symp. on Chem. of Bitumens, Rome, 1991. [25.] G.N. King, H.W. King and B. Brule, "Preparation and evaluation of polymer-modified asphalt binders," Proc. of ACS Rubber Division, Las Vegas, 1990. [26.] Personal communication, Ledo Quilici, Nevada Department of Transportation. [27.] M.G. Bouldin, "Results of repetitive creep experiments carried out on polymer modified hot mix asphalts used on the Mesquite Nevada job," Shell Development Co. Technical Bulletin, Houston, 1991. [28.] C.T. Chari, B.E. Ruth, M. Tia and G.C. Page, "Evaluation of age hardening of asphalts and mixtures," Association of Asphalt Paving Technologists Proc. AAPT 59 (1990), 176-239. [29.] C.J. Glover, R.R. Davison, S.M. Ghoreishi, H.B. Jenison and J.A. Bullin, "Evaluation of oven simulation of hot-mix aging by an FT-IR pellet procedure and other methods," TRB Record 880421, Washington, D. C., 1989. [30.] C.A. Bell, "Summary Report on aging of asphalt-aggregate systems," SHRP-A/1R-89-004, Washington, D.C., 1989. [31.] G.R. Kemp and N.H. Predoehl, "A comparison of field and laboratory environments on asphalt durability," Association of Asphalt Paving Technologists Proc. AAPT 50 (1981). [32.] M.G. Bouldin and M.A. Berggren, "The use of PBA materials in severe climates: Predictive tests," Shell Development Co. Technical Bulletin, Houston, 1991.