TECHNICAL NOTE

 

Determination of the continuous grade of a binder in the South African performance grade binder specification

 

 

S J Bredenhann; G M Rowe; P A Myburgh

Correspondence

 

 

---

ABSTRACT

A Performance Grade (PG) Bitumen Specification, SATS-3208 (SATS 2019), was developed in South Africa during the period up to 2019. Since that time a period of implementation has begun on selected projects in South Africa. The implementation activities have included testing of data in accordance with the final anticipated standards. Formulation of a national standard is in the final stages and is expected to be published towards the end of 2024. The PG grade is specified using a high temperature number and a minimum temperature number that indicate the climate range within which a bitumen would be expected to perform. For example, a PG64-22 has a high temperature of 64°C and a minimum temperature of -22°C, from which intermediate and low operating temperatures are derived. A material meeting this grade will always have a performance range which is, ideally, just bigger than the specification grade and this is typically referenced as the continuous grade. A further definition is the difference between the two continuous grade numbers, referenced as the "Useful Temperature Interval" or UTI. As part of this implementation, it is required to determine the continuous grade for bitumen, including for modified binders, to ensure compliance with the specification. The UTI must be equal to or greater than 80°C. This technical note has been developed to demonstrate correct procedure to determine the continuous grade and UTI.

Keywords: PG grade, bitumen, binder performance, fatigue, rutting, ageing ratio, continuous grading temperature

---

 

 

INTRODUCTION

The Performance Grade (PG) Bitumen Specification in South Africa (SA) was discussed by Bredenhann et al (2019) and can be downloaded from https://saice.org.za/journal/and is defined in SATS-3208 (SATS 2019) - a technical specification that will be followed by the SANS 4001:BT10 final standard specification (this specification is in the final stages of completion).

Each binder grade is defined by a high and low temperature. It is desirable that a binder must be sufficiently stiff at high temperatures to resist permanent deformation (rutting) and sufficiently flexible at low temperatures to resist low temperature cracking. At intermediate levels the stiffness must be low enough not to be brittle, and with adequate relaxation properties to resist fatigue cracking. The maximum pavement design temperature T_max is the maximum annual seven-day average temperature at the 97.5% confidence level in the asphalt layer at a depth of 20 mm. If a base layer is considered, the depth is taken as 20 mm into the base layer plus the asphalt cover on the base layer. The minimum grading temperature is the maximum pavement design temperature less 80°C, i.e. T_min = [T_max - 80]°C. In effect, a minimum Useful Temperature Interval (UTI) of 80°C ensures optimal performance at both intermediate and low temperatures. At the intermediate temperatures it is important that the binder has good relaxation properties so that the material is not susceptible to cracking. This is particularly important when binder is used in a surface course mix which is more at risk to ageing.

Fatigue in the SA PG specification is controlled by the minimum UTI together with an Ageing Ratio (AR) to limit stiffness and prevent brittleness after ageing and thus fatigue cracking at operational intermediate temperatures.

Permanent deformation is controlled with the non-recoverable strain (J_NR) at a creep stress of 3.2 kPa. The J_nr is determined with the Multiple Stress Creep and Recovery (MSCR) test as described in ASTM-D7175 (ASTM 2015).

Cracking, either fatigue or low temperature, is controlled with the minimum UTI, the AR and the ΔT_C, defined as T_c,S - T_c,m where the critical temperatures T_c,S and T_c,m are defined where stiffness S(60) = 300 MPa and m(60) = 0.3, as per ASTM-D7405 (ASTM 2020).

In the United States of America (USA) the rutting parameter implemented in 1994 was the |G*|/si nδ. This parameter has not been included in the SA PG specification, as it has been shown that |G*|/sinδ does not correlate well with field performance for modified binders and is slowly being replaced by use of the J_nr. G* and 5 are, however, required to be reported for an unaged binder, as it is required for the calculation of the Ageing Ratio.

It is the purpose of this technical note to describe the correct procedure to determine the continuous grade of a binder. Such a continuous grade is based on the maximum/minimum temperatures where the binder value corresponds with the prescribed minimum/maximum specification requirement. The procedure for determining continuous grades is described in ASTM-D7643 (ASTM 2022).

 

STRATEGIC HIGHWAY RESEARCH PROJECT (SHRP) HIGH TEMPERATURE PARAMETER AND MSCR UPDATE

Gundla et al (2020) discuss the relationship between |G*|/sinδ and J_nr and it is shown that |G*|/sinδ = 7.997 • (J_NR)^-0.85 with R² = 0.994. This relationship is generally valid for non-modified bitumen, but for modified binders this relationship does not apply. The work conducted by Gundla et al (2020) developed following analysis of earlier work by Kaloush et al (2019). Figure 1 shows the G*/sinδ and J_nr relationship graphically (the graph is reproduced with permission from data shared with the authors).

 

 

Figure 1 shows that a direct relationship exists between |G*|/sinδ and J_nr for non-modified binders. However, in the case of modified binders no such relationship was found. This supports the concept of grade bumping (i.e. specifying one to two grades higher than required by the environment) which has been used since the mid-1990s. However, as research was conducted with the MSCR and modified binders, it became clear that the concept of grade bumping, as applied with non-modified binders, cannot be applied in the same simplistic manner. With the MSCR test the same specification requirement is applied but at different values of J_NR specified at the given pavement temperature. This is the approach followed in the SA PG specification. Unfortunately, standard programs on the DSR equipment provide only the use of the USA standard of |G*|/sinδ, but this standard is not in the SA PG specification, so if used, the binders would be classified incorrectly. The correct approach to apply with the SA PG specification is discussed in this technical note.

Six samples of South African pen-grade bitumen were added to the database used by Gundla et al (2020) and are shown in Figure 2. Corresponding |G*|/sinδ and J_nr values for 70/100, 50/70 and 10/20 bitumen were included at 50-70°C. Only values |G*|/sinδ < 20 and J_nr < 4.5 were included to keep the scale of the two graphs the same.

 

 

It is deduced from Figure 2 that unmodified bitumen from South Africa follows the same trend as those tested in the mentioned study in the USA.

 

THE SA PG REQUIREMENT FOR PERMANENT DEFORMATION

In Table 1 the nomenclature used to describe the severity of the traffic loading is summarised.

 

 

DETERMINATION OF HIGH TEMPERATURE CONTINUOUS GRADE

Four bitumen types are compared in Figure 3, in terms of the SANS 4001:BT1 classification that is familiar to the industry - a 50/70 bitumen, a 70/100 bitumen, a 10/20 bitumen (all refined in local refineries) and a very hard 10/20 bitumen (imported). Test results from three laboratories are shown in Figure 3 to demonstrate firstly the variability in test results, and secondly the extent to which hard binder (the 10/20 PEN) test temperatures must be increased to obtain results.

 

 

The continuous grading temperatures and continuous grade for a binder graded in accordance with the requirements specified in ASTM D6373 are based on the critical temperatures where J_nr = 4.5 kPa^-1, as demonstrated in Figure 3 - simply put, the exact temperatures at which J_nr = 4.5 kPa^-1 are referred to as "true" high temperatures. The interpolation of the continuous temperature is made by considering the log of the J_nr versus temperature expressed on a linear scale. The continuous grade is"...a grade defined by the estimated upper and lower continuous grading temperatures while the continuous grading temperature Tc,n is the estimated temperatures at which the properties of an asphalt binder are equal to the specification requirement..." (ASTM-D7643) (ASTM 2022). The specification then refers to D6373 that will be SATS-3208 (SATS 2019) in the South African context where high temperature is defined with the MSCR test and low temperature with the BBR test.

PG grades are classified in 6°C intervals based on the finding that, for non-modified bitumen, for every 6°C increase/decrease the viscosity (therefore also shear modulus G*) will be halved/doubled. Continuous grades are rounded to the lower grade temperature.

The MSCR test is a destructive test, therefore a new binder sample is required for every temperature tested.

From Figure 3. it is deduced that continuous grades for the four bitumens are as shown in Table 2.

 

 

PG grades as shown in Table 2 are typical of bitumen produced in South African refineries, except for the imported 10/20 shown with a one grade higher T_max than the locally produced bitumen. It should also be noted from Table 2 that the SA 10/20 bitumen continuous grade value was determined through extrapolation, as not enough MSCR tests were done. Extrapolation should preferably not be done; interpolation is more accurate, and it should be ensured that an adequate number of samples are prepared to do MSCR at high enough temperatures to ensure interpolation.

 

DETERMINATION OF LOW TEMPERATURE CONTINUOUS GRADE

Continuous low temperature grades are determined from BBR tests in accordance with ASTM-D6448 (ASTM 2007). The low temperature continuous grade is developed in a similar manner to that for J_nr with the exception that two parameters are considered rather than one. At a loading time of 60 seconds in the bending beam rheometer (BBR) the stiffness "S" and the magnitude of the slope of the log stiffness versus log time curve (m) are used in the interpolation. The stiffness is expressed on a log basis, while m is plotted on a linear basis against temperature. The highest (least negative) of the two numbers is considered to be the low temperature grade. Determination of the low temperature continuous grade is done in accordance with ASTM-D7643 (ASTM 2022) applying Equations 1 and 2 to determine the critical temperatures. The parameter ATc is also calculated, from the equations (Equation 3) as follows:

Where:

T_{c, S} = temperature at which the stiffness is 300 MPa

T_c,m = temperature at which the m-value is 0.300

T₁, T₂ = two adjacent specification grading temperatures

S₁, S₂ = stiffness measurements at T₁, T₂ such that one value passes the specification requirement and one fails the requirement

m₁, m₂ = m-values at T₁, T₂ such that one value passes the specification requirement, and one fails the requirement.

For calculation purposes T₁ may be designated as the upper or lower temperature, as long as the corresponding specification test result (S or m) is used. It is good practice to use the S₁ as the value lower than 300 MPa and then select the corresponding T₁ temperature for consistency, and similar for the m values. Critical low temperature determination is illustrated in Figure 4 with:

 

 

Critical temperatures for all tested bitumens are shown in Table 3. The test requirement ΔTc = T_c,S - T_c,m is shown in Table 3, although it is not required to determine the continuous low temperature of a bitumen.

 

 

Critical temperatures T_c,s and T_c,m are reported at the test temperature that is 10°C higher than the actual temperature. All BBR tests are done at a temperature 10°C higher than the specified minimum temperature, e.g. for PG58-22 the isotherms are determined at -12°C to determine S(60) and m(60). The reason for this is that the original researchers who developed this test method considered the performance at two-hours loading time (7 200 seconds). This test time was considered too long and some calculations were performed, considering the time-temperature superposition princple, for typical binders in use to determine equivalencies to other temperatures. It was decided that testing at a loading time of 60 seconds and a ten-degree temperature shift provided the same test results based on the analysis.

 

FINAL CONTINUOUS GRADING

Finally, the high and low binder grades evaluated herein are reported in Table 4, indicating the final critical temperatures. The low temperature continuous grade reported in this table is the minimum of the values reported in Table 3 (Tc,S or Tc,m) with a correction of -10°C to arrive at the grade temperature.

 

 

Table 4 shows that all bitumen reported complies with the UTI > 80°C requirement.

As a rule, the 10/20 bitumens have high UTI values nearly meeting the expectations for a polymer modified binder. The reasons for this phenomenon are beyond the scope of this technical note.

 

ACCURACY OF MEASUREMENT

Accuracy of testing is very important; the dataset shown in Figure 4 is an example of a dataset that is below standard and was included to demonstrate the importance of accuracy. Data quality can be improved by duplicate tests. The S(60) versus temperature and m(60) versus temperature should be a smooth function. From inspection it can be deduced that the -12°C datapoint is clearly problematic. Accuracy can be improved by preparing samples at the same time and test within an acceptable timeframe to avoid secondary influences such as steric hardening, especially for BBR tests that are done at low temperatures. As mentioned already, critical temperatures must be determined through interpolation, not extrapolation.

 

CONCLUSION

Continuous performance grades should be determined from requirements in the South African PG Specification SATS 3208 (SATS 2019) which is soon to be superseded by SANS 4001:BT10. High continuous grade is determined from RTFO-aged bitumen utilising the J_nr parameter as determined in the MSCR test, while the low temperature continuous grade is determined from the BBR tests, the higher critical temperature determined from S(60) = 300 MPa or m(60) = 0.300.

 

REFERENCES

ASTM 2007. ASTM-D6448 2008. Standard Test Method for Determining the Flexural Creep Stiffness of Asphalt Binder using the Bending Beam Rheometer (BBR). West Conshohocken, PA: ASTM.         [ Links ]

ASTM 2015. ASTM-D7175 2015. Standard Test Method for Determining the Rheological Properties of Asphalt Binder using a Dynamic Shear Rheometer. West Conshohocken, PA: ASTM        [ Links ]

ASTM 2020. ASTM-D7405 2020. Standard Test Method for Multiple Stress Creep and Recovery (MSCR) of Asphalt Binder using a Dynamic Shear Rheometer. West Conshohocken, PA: ASTM International.         [ Links ]

ASTM 2022. ASTM-D7643 2022. Standard Practice for Determining the Continuous Grading Temperatures and Continuous Grades for PG Graded Asphalt Binders. West Conshohocken, PA: ASTM International.         [ Links ]

Bredenhann, S J, Myburgh, PA, Jenkins, K J, O'Connell, J S, Rowe, GM & D'Angelo, J 2019. Implementation of a performance-grade bitumen specification in South Africa. Journal of the South African Institution of Civil Engineering, 61(3): 20-31.         [ Links ]

Gundla, A, Salim, R, Underwood, B S & Kaloush, K 2020. Implementation of the AASHTO M332 Specification: A case study. Transportation Research Record, 2674(9): 959-971.         [ Links ]

Kaloush, K E, Underwood, B S, Salim, R & Gundla, A 2019. Evaluation of MSCR testing for adoption in ADOT asphalt binder specifications, Phoenix, AR: Federal Highway Administration.         [ Links ]

SATS (South African Technical Specification) 2019. SATS 3208 2019. Performance Grade (PG) Specifications for Bitumen in South Africa, Pretoria: South African Bureau of Standards.         [ Links ]

 

 

Correspondence:
S J Bredenhann
Naidu Consulting (Pty) Ltd 8 Gardner Williams Avenue Paardevlei Rising Unit 101 Somerset West 7130 South Africa
E: steph.bredenhann@naiduconsulting.com / steph@enpave.net

G M Rowe
President Abatech Inc PO Box 356 Blooming Glen PA 18911, United States of America
E: growe@abatech.com

P A Myburgh
House No 2, Meerenbosch Retirement Village
68 Ealcon Street, D'Urbanvale, Cape Town 7550, South Africa
E: pmyburgh@ffg.net

 

 

 

STEPH BREDENHANN (PrEng, PrCPM, FSAICE, HonSAT) graduated from Stellenbosch University in 1977 with a degree in Civil Engineering and completed a Master's degree in 2000. He started working at the Sishen Iron Ore Mine, but later joined the Kimberley City Council before taking up employmentasa directorwith consulting engineering firm Entech Consultants in Stellenbosch. He then joined WSP Consulting Engineers as a regional director before moving to Goba Consulting Engineers as technical director, and eventually to SANRAL (South African National Roads Agency Limited) as its western region project manager for research. He specialises in pavement and materials engineering, and is currently principal specialist at Naidu Consulting.

 

 

DR GEOEEREY ROWE (CEng, PE, FIAT, FCIHT, MIMMM, MASCE) has been working with asphalt materials for 43 years and is currently the president of Abatech, providing consultancy services on projects around the world. He has implemented significant technology for pavements, including pavement structural design, analysis of EWD data, materials property analysis and rheology. He is a past president of the Association of Asphalt Paving Technologists, and serves on many other groups advising on research that impacts specifications in North America, Europe, Asia and Africa.

 

 

PIETER MYBURGH (ESAICE, ESAT) graduated in 1965 from the University of Cape Town with a BSc in Civil Engineering, following which he was employed by the Provincial Government of the Cape Province as engineer in their Roads Department. Erom 1982 he was Chief Executive Officer of the Southern African Bitumen Association 06. Since then he has been acting as a private specialist consultant in pavement and materials engineering.